Subscribe to the PwC Newsletter

Join the community, trending research, graphcast: learning skillful medium-range global weather forecasting.

research paper machine learning

Global medium-range weather forecasting is critical to decision-making across many social and economic domains.

LCM-LoRA: A Universal Stable-Diffusion Acceleration Module

research paper machine learning

Latent Consistency Models (LCMs) have achieved impressive performance in accelerating text-to-image generative tasks, producing high-quality images with minimal inference steps.

research paper machine learning

Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models

qwenlm/qwen-audio • 14 Nov 2023

Recently, instruction-following audio-language models have received broad attention for audio interaction with humans.

S-LoRA: Serving Thousands of Concurrent LoRA Adapters

To capitalize on these opportunities, we present S-LoRA, a system designed for the scalable serving of many LoRA adapters.

Monkey: Image Resolution and Text Label Are Important Things for Large Multi-modal Models

yuliang-liu/monkey • 11 Nov 2023

Large Multimodal Models have demonstrated impressive capabilities in understanding general vision-language tasks.

research paper machine learning

The Rise and Potential of Large Language Model Based Agents: A Survey

woooodyy/llm-agent-paper-list • 14 Sep 2023

Many efforts have been made to develop intelligent agents, but they mainly focus on advancement in algorithms or training strategies to enhance specific capabilities or performance on particular tasks.

research paper machine learning

A Survey on Language Models for Code

codefuse-ai/awesome-code-llm • 14 Nov 2023

In this work we systematically review the recent advancements in code processing with language models, covering 50+ models, 30+ evaluation tasks, and 500 related works.

The Song Describer Dataset: a Corpus of Audio Captions for Music-and-Language Evaluation

We introduce the Song Describer dataset (SDD), a new crowdsourced corpus of high-quality audio-caption pairs, designed for the evaluation of music-and-language models.

JaxMARL: Multi-Agent RL Environments in JAX

This not only enables GPU acceleration, but also provides a more flexible MARL environment, unlocking the potential for self-play, meta-learning, and other future applications in MARL.

research paper machine learning

Plum: Prompt Learning using Metaheuristic

Since the emergence of large language models, prompt learning has become a popular method for optimizing and customizing these models.

U.S. flag

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

  • Publications
  • Account settings
  • Advanced Search
  • Journal List
  • Springer Nature - PMC COVID-19 Collection

Logo of phenaturepg

Machine Learning: Algorithms, Real-World Applications and Research Directions

Iqbal h. sarker.

1 Swinburne University of Technology, Melbourne, VIC 3122 Australia

2 Department of Computer Science and Engineering, Chittagong University of Engineering & Technology, 4349 Chattogram, Bangladesh

In the current age of the Fourth Industrial Revolution (4 IR or Industry 4.0), the digital world has a wealth of data, such as Internet of Things (IoT) data, cybersecurity data, mobile data, business data, social media data, health data, etc. To intelligently analyze these data and develop the corresponding smart and automated  applications, the knowledge of artificial intelligence (AI), particularly, machine learning (ML) is the key. Various types of machine learning algorithms such as supervised, unsupervised, semi-supervised, and reinforcement learning exist in the area. Besides, the deep learning , which is part of a broader family of machine learning methods, can intelligently analyze the data on a large scale. In this paper, we present a comprehensive view on these machine learning algorithms that can be applied to enhance the intelligence and the capabilities of an application. Thus, this study’s key contribution is explaining the principles of different machine learning techniques and their applicability in various real-world application domains, such as cybersecurity systems, smart cities, healthcare, e-commerce, agriculture, and many more. We also highlight the challenges and potential research directions based on our study. Overall, this paper aims to serve as a reference point for both academia and industry professionals as well as for decision-makers in various real-world situations and application areas, particularly from the technical point of view.


We live in the age of data, where everything around us is connected to a data source, and everything in our lives is digitally recorded [ 21 , 103 ]. For instance, the current electronic world has a wealth of various kinds of data, such as the Internet of Things (IoT) data, cybersecurity data, smart city data, business data, smartphone data, social media data, health data, COVID-19 data, and many more. The data can be structured, semi-structured, or unstructured, discussed briefly in Sect. “ Types of Real-World Data and Machine Learning Techniques ”, which is increasing day-by-day. Extracting insights from these data can be used to build various intelligent applications in the relevant domains. For instance, to build a data-driven automated and intelligent cybersecurity system, the relevant cybersecurity data can be used [ 105 ]; to build personalized context-aware smart mobile applications, the relevant mobile data can be used [ 103 ], and so on. Thus, the data management tools and techniques having the capability of extracting insights or useful knowledge from the data in a timely and intelligent way is urgently needed, on which the real-world applications are based.

Artificial intelligence (AI), particularly, machine learning (ML) have grown rapidly in recent years in the context of data analysis and computing that typically allows the applications to function in an intelligent manner [ 95 ]. ML usually provides systems with the ability to learn and enhance from experience automatically without being specifically programmed and is generally referred to as the most popular latest technologies in the fourth industrial revolution (4 IR or Industry 4.0) [ 103 , 105 ]. “Industry 4.0” [ 114 ] is typically the ongoing automation of conventional manufacturing and industrial practices, including exploratory data processing, using new smart technologies such as machine learning automation. Thus, to intelligently analyze these data and to develop the corresponding real-world applications, machine learning algorithms is the key. The learning algorithms can be categorized into four major types, such as supervised, unsupervised, semi-supervised, and reinforcement learning in the area [ 75 ], discussed briefly in Sect. “ Types of Real-World Data and Machine Learning Techniques ”. The popularity of these approaches to learning is increasing day-by-day, which is shown in Fig. ​ Fig.1, 1 , based on data collected from Google Trends [ 4 ] over the last five years. The x - axis of the figure indicates the specific dates and the corresponding popularity score within the range of 0 ( m i n i m u m ) to 100 ( m a x i m u m ) has been shown in y - axis . According to Fig. ​ Fig.1, 1 , the popularity indication values for these learning types are low in 2015 and are increasing day by day. These statistics motivate us to study on machine learning in this paper, which can play an important role in the real-world through Industry 4.0 automation.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig1_HTML.jpg

The worldwide popularity score of various types of ML algorithms (supervised, unsupervised, semi-supervised, and reinforcement) in a range of 0 (min) to 100 (max) over time where x-axis represents the timestamp information and y-axis represents the corresponding score

In general, the effectiveness and the efficiency of a machine learning solution depend on the nature and characteristics of data and the performance of the learning algorithms . In the area of machine learning algorithms, classification analysis, regression, data clustering, feature engineering and dimensionality reduction, association rule learning, or reinforcement learning techniques exist to effectively build data-driven systems [ 41 , 125 ]. Besides, deep learning originated from the artificial neural network that can be used to intelligently analyze data, which is known as part of a wider family of machine learning approaches [ 96 ]. Thus, selecting a proper learning algorithm that is suitable for the target application in a particular domain is challenging. The reason is that the purpose of different learning algorithms is different, even the outcome of different learning algorithms in a similar category may vary depending on the data characteristics [ 106 ]. Thus, it is important to understand the principles of various machine learning algorithms and their applicability to apply in various real-world application areas, such as IoT systems, cybersecurity services, business and recommendation systems, smart cities, healthcare and COVID-19, context-aware systems, sustainable agriculture, and many more that are explained briefly in Sect. “ Applications of Machine Learning ”.

Based on the importance and potentiality of “Machine Learning” to analyze the data mentioned above, in this paper, we provide a comprehensive view on various types of machine learning algorithms that can be applied to enhance the intelligence and the capabilities of an application. Thus, the key contribution of this study is explaining the principles and potentiality of different machine learning techniques, and their applicability in various real-world application areas mentioned earlier. The purpose of this paper is, therefore, to provide a basic guide for those academia and industry people who want to study, research, and develop data-driven automated and intelligent systems in the relevant areas based on machine learning techniques.

The key contributions of this paper are listed as follows:

  • To define the scope of our study by taking into account the nature and characteristics of various types of real-world data and the capabilities of various learning techniques.
  • To provide a comprehensive view on machine learning algorithms that can be applied to enhance the intelligence and capabilities of a data-driven application.
  • To discuss the applicability of machine learning-based solutions in various real-world application domains.
  • To highlight and summarize the potential research directions within the scope of our study for intelligent data analysis and services.

The rest of the paper is organized as follows. The next section presents the types of data and machine learning algorithms in a broader sense and defines the scope of our study. We briefly discuss and explain different machine learning algorithms in the subsequent section followed by which various real-world application areas based on machine learning algorithms are discussed and summarized. In the penultimate section, we highlight several research issues and potential future directions, and the final section concludes this paper.

Types of Real-World Data and Machine Learning Techniques

Machine learning algorithms typically consume and process data to learn the related patterns about individuals, business processes, transactions, events, and so on. In the following, we discuss various types of real-world data as well as categories of machine learning algorithms.

Types of Real-World Data

Usually, the availability of data is considered as the key to construct a machine learning model or data-driven real-world systems [ 103 , 105 ]. Data can be of various forms, such as structured, semi-structured, or unstructured [ 41 , 72 ]. Besides, the “metadata” is another type that typically represents data about the data. In the following, we briefly discuss these types of data.

  • Structured: It has a well-defined structure, conforms to a data model following a standard order, which is highly organized and easily accessed, and used by an entity or a computer program. In well-defined schemes, such as relational databases, structured data are typically stored, i.e., in a tabular format. For instance, names, dates, addresses, credit card numbers, stock information, geolocation, etc. are examples of structured data.
  • Unstructured: On the other hand, there is no pre-defined format or organization for unstructured data, making it much more difficult to capture, process, and analyze, mostly containing text and multimedia material. For example, sensor data, emails, blog entries, wikis, and word processing documents, PDF files, audio files, videos, images, presentations, web pages, and many other types of business documents can be considered as unstructured data.
  • Semi-structured: Semi-structured data are not stored in a relational database like the structured data mentioned above, but it does have certain organizational properties that make it easier to analyze. HTML, XML, JSON documents, NoSQL databases, etc., are some examples of semi-structured data.
  • Metadata: It is not the normal form of data, but “data about data”. The primary difference between “data” and “metadata” is that data are simply the material that can classify, measure, or even document something relative to an organization’s data properties. On the other hand, metadata describes the relevant data information, giving it more significance for data users. A basic example of a document’s metadata might be the author, file size, date generated by the document, keywords to define the document, etc.

In the area of machine learning and data science, researchers use various widely used datasets for different purposes. These are, for example, cybersecurity datasets such as NSL-KDD [ 119 ], UNSW-NB15 [ 76 ], ISCX’12 [ 1 ], CIC-DDoS2019 [ 2 ], Bot-IoT [ 59 ], etc., smartphone datasets such as phone call logs [ 84 , 101 ], SMS Log [ 29 ], mobile application usages logs [ 137 ] [ 117 ], mobile phone notification logs [ 73 ] etc., IoT data [ 16 , 57 , 62 ], agriculture and e-commerce data [ 120 , 138 ], health data such as heart disease [ 92 ], diabetes mellitus [ 83 , 134 ], COVID-19 [ 43 , 74 ], etc., and many more in various application domains. The data can be in different types discussed above, which may vary from application to application in the real world. To analyze such data in a particular problem domain, and to extract the insights or useful knowledge from the data for building the real-world intelligent applications, different types of machine learning techniques can be used according to their learning capabilities, which is discussed in the following.

Types of Machine Learning Techniques

Machine Learning algorithms are mainly divided into four categories: Supervised learning, Unsupervised learning, Semi-supervised learning, and Reinforcement learning [ 75 ], as shown in Fig. ​ Fig.2. 2 . In the following, we briefly discuss each type of learning technique with the scope of their applicability to solve real-world problems.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig2_HTML.jpg

Various types of machine learning techniques

  • Supervised: Supervised learning is typically the task of machine learning to learn a function that maps an input to an output based on sample input-output pairs [ 41 ]. It uses labeled training data and a collection of training examples to infer a function. Supervised learning is carried out when certain goals are identified to be accomplished from a certain set of inputs [ 105 ], i.e., a task-driven approach . The most common supervised tasks are “classification” that separates the data, and “regression” that fits the data. For instance, predicting the class label or sentiment of a piece of text, like a tweet or a product review, i.e., text classification, is an example of supervised learning.
  • Unsupervised: Unsupervised learning analyzes unlabeled datasets without the need for human interference, i.e., a data-driven process [ 41 ]. This is widely used for extracting generative features, identifying meaningful trends and structures, groupings in results, and exploratory purposes. The most common unsupervised learning tasks are clustering, density estimation, feature learning, dimensionality reduction, finding association rules, anomaly detection, etc.
  • Semi-supervised: Semi-supervised learning can be defined as a hybridization of the above-mentioned supervised and unsupervised methods, as it operates on both labeled and unlabeled data [ 41 , 105 ]. Thus, it falls between learning “without supervision” and learning “with supervision”. In the real world, labeled data could be rare in several contexts, and unlabeled data are numerous, where semi-supervised learning is useful [ 75 ]. The ultimate goal of a semi-supervised learning model is to provide a better outcome for prediction than that produced using the labeled data alone from the model. Some application areas where semi-supervised learning is used include machine translation, fraud detection, labeling data and text classification.
  • Reinforcement: Reinforcement learning is a type of machine learning algorithm that enables software agents and machines to automatically evaluate the optimal behavior in a particular context or environment to improve its efficiency [ 52 ], i.e., an environment-driven approach . This type of learning is based on reward or penalty, and its ultimate goal is to use insights obtained from environmental activists to take action to increase the reward or minimize the risk [ 75 ]. It is a powerful tool for training AI models that can help increase automation or optimize the operational efficiency of sophisticated systems such as robotics, autonomous driving tasks, manufacturing and supply chain logistics, however, not preferable to use it for solving the basic or straightforward problems.

Thus, to build effective models in various application areas different types of machine learning techniques can play a significant role according to their learning capabilities, depending on the nature of the data discussed earlier, and the target outcome. In Table ​ Table1, 1 , we summarize various types of machine learning techniques with examples. In the following, we provide a comprehensive view of machine learning algorithms that can be applied to enhance the intelligence and capabilities of a data-driven application.

Various types of machine learning techniques with examples

Machine Learning Tasks and Algorithms

In this section, we discuss various machine learning algorithms that include classification analysis, regression analysis, data clustering, association rule learning, feature engineering for dimensionality reduction, as well as deep learning methods. A general structure of a machine learning-based predictive model has been shown in Fig. ​ Fig.3, 3 , where the model is trained from historical data in phase 1 and the outcome is generated in phase 2 for the new test data.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig3_HTML.jpg

A general structure of a machine learning based predictive model considering both the training and testing phase

Classification Analysis

Classification is regarded as a supervised learning method in machine learning, referring to a problem of predictive modeling as well, where a class label is predicted for a given example [ 41 ]. Mathematically, it maps a function ( f ) from input variables ( X ) to output variables ( Y ) as target, label or categories. To predict the class of given data points, it can be carried out on structured or unstructured data. For example, spam detection such as “spam” and “not spam” in email service providers can be a classification problem. In the following, we summarize the common classification problems.

  • Binary classification: It refers to the classification tasks having two class labels such as “true and false” or “yes and no” [ 41 ]. In such binary classification tasks, one class could be the normal state, while the abnormal state could be another class. For instance, “cancer not detected” is the normal state of a task that involves a medical test, and “cancer detected” could be considered as the abnormal state. Similarly, “spam” and “not spam” in the above example of email service providers are considered as binary classification.
  • Multiclass classification: Traditionally, this refers to those classification tasks having more than two class labels [ 41 ]. The multiclass classification does not have the principle of normal and abnormal outcomes, unlike binary classification tasks. Instead, within a range of specified classes, examples are classified as belonging to one. For example, it can be a multiclass classification task to classify various types of network attacks in the NSL-KDD [ 119 ] dataset, where the attack categories are classified into four class labels, such as DoS (Denial of Service Attack), U2R (User to Root Attack), R2L (Root to Local Attack), and Probing Attack.
  • Multi-label classification: In machine learning, multi-label classification is an important consideration where an example is associated with several classes or labels. Thus, it is a generalization of multiclass classification, where the classes involved in the problem are hierarchically structured, and each example may simultaneously belong to more than one class in each hierarchical level, e.g., multi-level text classification. For instance, Google news can be presented under the categories of a “city name”, “technology”, or “latest news”, etc. Multi-label classification includes advanced machine learning algorithms that support predicting various mutually non-exclusive classes or labels, unlike traditional classification tasks where class labels are mutually exclusive [ 82 ].

Many classification algorithms have been proposed in the machine learning and data science literature [ 41 , 125 ]. In the following, we summarize the most common and popular methods that are used widely in various application areas.

  • Naive Bayes (NB): The naive Bayes algorithm is based on the Bayes’ theorem with the assumption of independence between each pair of features [ 51 ]. It works well and can be used for both binary and multi-class categories in many real-world situations, such as document or text classification, spam filtering, etc. To effectively classify the noisy instances in the data and to construct a robust prediction model, the NB classifier can be used [ 94 ]. The key benefit is that, compared to more sophisticated approaches, it needs a small amount of training data to estimate the necessary parameters and quickly [ 82 ]. However, its performance may affect due to its strong assumptions on features independence. Gaussian, Multinomial, Complement, Bernoulli, and Categorical are the common variants of NB classifier [ 82 ].
  • Linear Discriminant Analysis (LDA): Linear Discriminant Analysis (LDA) is a linear decision boundary classifier created by fitting class conditional densities to data and applying Bayes’ rule [ 51 , 82 ]. This method is also known as a generalization of Fisher’s linear discriminant, which projects a given dataset into a lower-dimensional space, i.e., a reduction of dimensionality that minimizes the complexity of the model or reduces the resulting model’s computational costs. The standard LDA model usually suits each class with a Gaussian density, assuming that all classes share the same covariance matrix [ 82 ]. LDA is closely related to ANOVA (analysis of variance) and regression analysis, which seek to express one dependent variable as a linear combination of other features or measurements.
  • Logistic regression (LR): Another common probabilistic based statistical model used to solve classification issues in machine learning is Logistic Regression (LR) [ 64 ]. Logistic regression typically uses a logistic function to estimate the probabilities, which is also referred to as the mathematically defined sigmoid function in Eq. 1 . It can overfit high-dimensional datasets and works well when the dataset can be separated linearly. The regularization (L1 and L2) techniques [ 82 ] can be used to avoid over-fitting in such scenarios. The assumption of linearity between the dependent and independent variables is considered as a major drawback of Logistic Regression. It can be used for both classification and regression problems, but it is more commonly used for classification. g ( z ) = 1 1 + exp ( - z ) . 1
  • K-nearest neighbors (KNN): K-Nearest Neighbors (KNN) [ 9 ] is an “instance-based learning” or non-generalizing learning, also known as a “lazy learning” algorithm. It does not focus on constructing a general internal model; instead, it stores all instances corresponding to training data in n -dimensional space. KNN uses data and classifies new data points based on similarity measures (e.g., Euclidean distance function) [ 82 ]. Classification is computed from a simple majority vote of the k nearest neighbors of each point. It is quite robust to noisy training data, and accuracy depends on the data quality. The biggest issue with KNN is to choose the optimal number of neighbors to be considered. KNN can be used both for classification as well as regression.
  • Support vector machine (SVM): In machine learning, another common technique that can be used for classification, regression, or other tasks is a support vector machine (SVM) [ 56 ]. In high- or infinite-dimensional space, a support vector machine constructs a hyper-plane or set of hyper-planes. Intuitively, the hyper-plane, which has the greatest distance from the nearest training data points in any class, achieves a strong separation since, in general, the greater the margin, the lower the classifier’s generalization error. It is effective in high-dimensional spaces and can behave differently based on different mathematical functions known as the kernel. Linear, polynomial, radial basis function (RBF), sigmoid, etc., are the popular kernel functions used in SVM classifier [ 82 ]. However, when the data set contains more noise, such as overlapping target classes, SVM does not perform well.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig4_HTML.jpg

An example of a decision tree structure

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig5_HTML.jpg

An example of a random forest structure considering multiple decision trees

  • Adaptive Boosting (AdaBoost): Adaptive Boosting (AdaBoost) is an ensemble learning process that employs an iterative approach to improve poor classifiers by learning from their errors. This is developed by Yoav Freund et al. [ 35 ] and also known as “meta-learning”. Unlike the random forest that uses parallel ensembling, Adaboost uses “sequential ensembling”. It creates a powerful classifier by combining many poorly performing classifiers to obtain a good classifier of high accuracy. In that sense, AdaBoost is called an adaptive classifier by significantly improving the efficiency of the classifier, but in some instances, it can trigger overfits. AdaBoost is best used to boost the performance of decision trees, base estimator [ 82 ], on binary classification problems, however, is sensitive to noisy data and outliers.
  • Extreme gradient boosting (XGBoost): Gradient Boosting, like Random Forests [ 19 ] above, is an ensemble learning algorithm that generates a final model based on a series of individual models, typically decision trees. The gradient is used to minimize the loss function, similar to how neural networks [ 41 ] use gradient descent to optimize weights. Extreme Gradient Boosting (XGBoost) is a form of gradient boosting that takes more detailed approximations into account when determining the best model [ 82 ]. It computes second-order gradients of the loss function to minimize loss and advanced regularization (L1 and L2) [ 82 ], which reduces over-fitting, and improves model generalization and performance. XGBoost is fast to interpret and can handle large-sized datasets well.
  • Stochastic gradient descent (SGD): Stochastic gradient descent (SGD) [ 41 ] is an iterative method for optimizing an objective function with appropriate smoothness properties, where the word ‘stochastic’ refers to random probability. This reduces the computational burden, particularly in high-dimensional optimization problems, allowing for faster iterations in exchange for a lower convergence rate. A gradient is the slope of a function that calculates a variable’s degree of change in response to another variable’s changes. Mathematically, the Gradient Descent is a convex function whose output is a partial derivative of a set of its input parameters. Let, α is the learning rate, and J i is the training example cost of i th , then Eq. ( 4 ) represents the stochastic gradient descent weight update method at the j th iteration. In large-scale and sparse machine learning, SGD has been successfully applied to problems often encountered in text classification and natural language processing [ 82 ]. However, SGD is sensitive to feature scaling and needs a range of hyperparameters, such as the regularization parameter and the number of iterations. w j : = w j - α ∂ J i ∂ w j . 4
  • Rule-based classification : The term rule-based classification can be used to refer to any classification scheme that makes use of IF-THEN rules for class prediction. Several classification algorithms such as Zero-R [ 125 ], One-R [ 47 ], decision trees [ 87 , 88 ], DTNB [ 110 ], Ripple Down Rule learner (RIDOR) [ 125 ], Repeated Incremental Pruning to Produce Error Reduction (RIPPER) [ 126 ] exist with the ability of rule generation. The decision tree is one of the most common rule-based classification algorithms among these techniques because it has several advantages, such as being easier to interpret; the ability to handle high-dimensional data; simplicity and speed; good accuracy; and the capability to produce rules for human clear and understandable classification [ 127 ] [ 128 ]. The decision tree-based rules also provide significant accuracy in a prediction model for unseen test cases [ 106 ]. Since the rules are easily interpretable, these rule-based classifiers are often used to produce descriptive models that can describe a system including the entities and their relationships.

Regression Analysis

Regression analysis includes several methods of machine learning that allow to predict a continuous ( y ) result variable based on the value of one or more ( x ) predictor variables [ 41 ]. The most significant distinction between classification and regression is that classification predicts distinct class labels, while regression facilitates the prediction of a continuous quantity. Figure ​ Figure6 6 shows an example of how classification is different with regression models. Some overlaps are often found between the two types of machine learning algorithms. Regression models are now widely used in a variety of fields, including financial forecasting or prediction, cost estimation, trend analysis, marketing, time series estimation, drug response modeling, and many more. Some of the familiar types of regression algorithms are linear, polynomial, lasso and ridge regression, etc., which are explained briefly in the following.

  • Simple and multiple linear regression: This is one of the most popular ML modeling techniques as well as a well-known regression technique. In this technique, the dependent variable is continuous, the independent variable(s) can be continuous or discrete, and the form of the regression line is linear. Linear regression creates a relationship between the dependent variable ( Y ) and one or more independent variables ( X ) (also known as regression line) using the best fit straight line [ 41 ]. It is defined by the following equations: y = a + b x + e 5 y = a + b 1 x 1 + b 2 x 2 + ⋯ + b n x n + e , 6 where a is the intercept, b is the slope of the line, and e is the error term. This equation can be used to predict the value of the target variable based on the given predictor variable(s). Multiple linear regression is an extension of simple linear regression that allows two or more predictor variables to model a response variable, y, as a linear function [ 41 ] defined in Eq. 6 , whereas simple linear regression has only 1 independent variable, defined in Eq. 5 .
  • Polynomial regression: Polynomial regression is a form of regression analysis in which the relationship between the independent variable x and the dependent variable y is not linear, but is the polynomial degree of n th in x [ 82 ]. The equation for polynomial regression is also derived from linear regression (polynomial regression of degree 1) equation, which is defined as below: y = b 0 + b 1 x + b 2 x 2 + b 3 x 3 + ⋯ + b n x n + e . 7 Here, y is the predicted/target output, b 0 , b 1 , . . . b n are the regression coefficients, x is an independent/ input variable. In simple words, we can say that if data are not distributed linearly, instead it is n th degree of polynomial then we use polynomial regression to get desired output.
  • LASSO and ridge regression: LASSO and Ridge regression are well known as powerful techniques which are typically used for building learning models in presence of a large number of features, due to their capability to preventing over-fitting and reducing the complexity of the model. The LASSO (least absolute shrinkage and selection operator) regression model uses L 1 regularization technique [ 82 ] that uses shrinkage, which penalizes “absolute value of magnitude of coefficients” ( L 1 penalty). As a result, LASSO appears to render coefficients to absolute zero. Thus, LASSO regression aims to find the subset of predictors that minimizes the prediction error for a quantitative response variable. On the other hand, ridge regression uses L 2 regularization [ 82 ], which is the “squared magnitude of coefficients” ( L 2 penalty). Thus, ridge regression forces the weights to be small but never sets the coefficient value to zero, and does a non-sparse solution. Overall, LASSO regression is useful to obtain a subset of predictors by eliminating less important features, and ridge regression is useful when a data set has “multicollinearity” which refers to the predictors that are correlated with other predictors.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig6_HTML.jpg

Classification vs. regression. In classification the dotted line represents a linear boundary that separates the two classes; in regression, the dotted line models the linear relationship between the two variables

Cluster Analysis

Cluster analysis, also known as clustering, is an unsupervised machine learning technique for identifying and grouping related data points in large datasets without concern for the specific outcome. It does grouping a collection of objects in such a way that objects in the same category, called a cluster, are in some sense more similar to each other than objects in other groups [ 41 ]. It is often used as a data analysis technique to discover interesting trends or patterns in data, e.g., groups of consumers based on their behavior. In a broad range of application areas, such as cybersecurity, e-commerce, mobile data processing, health analytics, user modeling and behavioral analytics, clustering can be used. In the following, we briefly discuss and summarize various types of clustering methods.

  • Partitioning methods: Based on the features and similarities in the data, this clustering approach categorizes the data into multiple groups or clusters. The data scientists or analysts typically determine the number of clusters either dynamically or statically depending on the nature of the target applications, to produce for the methods of clustering. The most common clustering algorithms based on partitioning methods are K-means [ 69 ], K-Mediods [ 80 ], CLARA [ 55 ] etc.
  • Density-based methods: To identify distinct groups or clusters, it uses the concept that a cluster in the data space is a contiguous region of high point density isolated from other such clusters by contiguous regions of low point density. Points that are not part of a cluster are considered as noise. The typical clustering algorithms based on density are DBSCAN [ 32 ], OPTICS [ 12 ] etc. The density-based methods typically struggle with clusters of similar density and high dimensionality data.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig7_HTML.jpg

A graphical interpretation of the widely-used hierarchical clustering (Bottom-up and top-down) technique

  • Grid-based methods: To deal with massive datasets, grid-based clustering is especially suitable. To obtain clusters, the principle is first to summarize the dataset with a grid representation and then to combine grid cells. STING [ 122 ], CLIQUE [ 6 ], etc. are the standard algorithms of grid-based clustering.
  • Model-based methods: There are mainly two types of model-based clustering algorithms: one that uses statistical learning, and the other based on a method of neural network learning [ 130 ]. For instance, GMM [ 89 ] is an example of a statistical learning method, and SOM [ 22 ] [ 96 ] is an example of a neural network learning method.
  • Constraint-based methods: Constrained-based clustering is a semi-supervised approach to data clustering that uses constraints to incorporate domain knowledge. Application or user-oriented constraints are incorporated to perform the clustering. The typical algorithms of this kind of clustering are COP K-means [ 121 ], CMWK-Means [ 27 ], etc.

Many clustering algorithms have been proposed with the ability to grouping data in machine learning and data science literature [ 41 , 125 ]. In the following, we summarize the popular methods that are used widely in various application areas.

  • K-means clustering: K-means clustering [ 69 ] is a fast, robust, and simple algorithm that provides reliable results when data sets are well-separated from each other. The data points are allocated to a cluster in this algorithm in such a way that the amount of the squared distance between the data points and the centroid is as small as possible. In other words, the K-means algorithm identifies the k number of centroids and then assigns each data point to the nearest cluster while keeping the centroids as small as possible. Since it begins with a random selection of cluster centers, the results can be inconsistent. Since extreme values can easily affect a mean, the K-means clustering algorithm is sensitive to outliers. K-medoids clustering [ 91 ] is a variant of K-means that is more robust to noises and outliers.
  • Mean-shift clustering: Mean-shift clustering [ 37 ] is a nonparametric clustering technique that does not require prior knowledge of the number of clusters or constraints on cluster shape. Mean-shift clustering aims to discover “blobs” in a smooth distribution or density of samples [ 82 ]. It is a centroid-based algorithm that works by updating centroid candidates to be the mean of the points in a given region. To form the final set of centroids, these candidates are filtered in a post-processing stage to remove near-duplicates. Cluster analysis in computer vision and image processing are examples of application domains. Mean Shift has the disadvantage of being computationally expensive. Moreover, in cases of high dimension, where the number of clusters shifts abruptly, the mean-shift algorithm does not work well.
  • DBSCAN: Density-based spatial clustering of applications with noise (DBSCAN) [ 32 ] is a base algorithm for density-based clustering which is widely used in data mining and machine learning. This is known as a non-parametric density-based clustering technique for separating high-density clusters from low-density clusters that are used in model building. DBSCAN’s main idea is that a point belongs to a cluster if it is close to many points from that cluster. It can find clusters of various shapes and sizes in a vast volume of data that is noisy and contains outliers. DBSCAN, unlike k-means, does not require a priori specification of the number of clusters in the data and can find arbitrarily shaped clusters. Although k-means is much faster than DBSCAN, it is efficient at finding high-density regions and outliers, i.e., is robust to outliers.
  • GMM clustering: Gaussian mixture models (GMMs) are often used for data clustering, which is a distribution-based clustering algorithm. A Gaussian mixture model is a probabilistic model in which all the data points are produced by a mixture of a finite number of Gaussian distributions with unknown parameters [ 82 ]. To find the Gaussian parameters for each cluster, an optimization algorithm called expectation-maximization (EM) [ 82 ] can be used. EM is an iterative method that uses a statistical model to estimate the parameters. In contrast to k-means, Gaussian mixture models account for uncertainty and return the likelihood that a data point belongs to one of the k clusters. GMM clustering is more robust than k-means and works well even with non-linear data distributions.
  • Agglomerative hierarchical clustering: The most common method of hierarchical clustering used to group objects in clusters based on their similarity is agglomerative clustering. This technique uses a bottom-up approach, where each object is first treated as a singleton cluster by the algorithm. Following that, pairs of clusters are merged one by one until all clusters have been merged into a single large cluster containing all objects. The result is a dendrogram, which is a tree-based representation of the elements. Single linkage [ 115 ], Complete linkage [ 116 ], BOTS [ 102 ] etc. are some examples of such techniques. The main advantage of agglomerative hierarchical clustering over k-means is that the tree-structure hierarchy generated by agglomerative clustering is more informative than the unstructured collection of flat clusters returned by k-means, which can help to make better decisions in the relevant application areas.

Dimensionality Reduction and Feature Learning

In machine learning and data science, high-dimensional data processing is a challenging task for both researchers and application developers. Thus, dimensionality reduction which is an unsupervised learning technique, is important because it leads to better human interpretations, lower computational costs, and avoids overfitting and redundancy by simplifying models. Both the process of feature selection and feature extraction can be used for dimensionality reduction. The primary distinction between the selection and extraction of features is that the “feature selection” keeps a subset of the original features [ 97 ], while “feature extraction” creates brand new ones [ 98 ]. In the following, we briefly discuss these techniques.

  • Feature selection: The selection of features, also known as the selection of variables or attributes in the data, is the process of choosing a subset of unique features (variables, predictors) to use in building machine learning and data science model. It decreases a model’s complexity by eliminating the irrelevant or less important features and allows for faster training of machine learning algorithms. A right and optimal subset of the selected features in a problem domain is capable to minimize the overfitting problem through simplifying and generalizing the model as well as increases the model’s accuracy [ 97 ]. Thus, “feature selection” [ 66 , 99 ] is considered as one of the primary concepts in machine learning that greatly affects the effectiveness and efficiency of the target machine learning model. Chi-squared test, Analysis of variance (ANOVA) test, Pearson’s correlation coefficient, recursive feature elimination, are some popular techniques that can be used for feature selection.
  • Feature extraction: In a machine learning-based model or system, feature extraction techniques usually provide a better understanding of the data, a way to improve prediction accuracy, and to reduce computational cost or training time. The aim of “feature extraction” [ 66 , 99 ] is to reduce the number of features in a dataset by generating new ones from the existing ones and then discarding the original features. The majority of the information found in the original set of features can then be summarized using this new reduced set of features. For instance, principal components analysis (PCA) is often used as a dimensionality-reduction technique to extract a lower-dimensional space creating new brand components from the existing features in a dataset [ 98 ].

Many algorithms have been proposed to reduce data dimensions in the machine learning and data science literature [ 41 , 125 ]. In the following, we summarize the popular methods that are used widely in various application areas.

  • Variance threshold: A simple basic approach to feature selection is the variance threshold [ 82 ]. This excludes all features of low variance, i.e., all features whose variance does not exceed the threshold. It eliminates all zero-variance characteristics by default, i.e., characteristics that have the same value in all samples. This feature selection algorithm looks only at the ( X ) features, not the ( y ) outputs needed, and can, therefore, be used for unsupervised learning.
  • Pearson correlation: Pearson’s correlation is another method to understand a feature’s relation to the response variable and can be used for feature selection [ 99 ]. This method is also used for finding the association between the features in a dataset. The resulting value is [ - 1 , 1 ] , where - 1 means perfect negative correlation, + 1 means perfect positive correlation, and 0 means that the two variables do not have a linear correlation. If two random variables represent X and Y , then the correlation coefficient between X and Y is defined as [ 41 ] r ( X , Y ) = ∑ i = 1 n ( X i - X ¯ ) ( Y i - Y ¯ ) ∑ i = 1 n ( X i - X ¯ ) 2 ∑ i = 1 n ( Y i - Y ¯ ) 2 . 8
  • ANOVA: Analysis of variance (ANOVA) is a statistical tool used to verify the mean values of two or more groups that differ significantly from each other. ANOVA assumes a linear relationship between the variables and the target and the variables’ normal distribution. To statistically test the equality of means, the ANOVA method utilizes F tests. For feature selection, the results ‘ANOVA F value’ [ 82 ] of this test can be used where certain features independent of the goal variable can be omitted.
  • Chi square: The chi-square χ 2 [ 82 ] statistic is an estimate of the difference between the effects of a series of events or variables observed and expected frequencies. The magnitude of the difference between the real and observed values, the degrees of freedom, and the sample size depends on χ 2 . The chi-square χ 2 is commonly used for testing relationships between categorical variables. If O i represents observed value and E i represents expected value, then χ 2 = ∑ i = 1 n ( O i - E i ) 2 E i . 9
  • Recursive feature elimination (RFE): Recursive Feature Elimination (RFE) is a brute force approach to feature selection. RFE [ 82 ] fits the model and removes the weakest feature before it meets the specified number of features. Features are ranked by the coefficients or feature significance of the model. RFE aims to remove dependencies and collinearity in the model by recursively removing a small number of features per iteration.
  • Model-based selection: To reduce the dimensionality of the data, linear models penalized with the L 1 regularization can be used. Least absolute shrinkage and selection operator (Lasso) regression is a type of linear regression that has the property of shrinking some of the coefficients to zero [ 82 ]. Therefore, that feature can be removed from the model. Thus, the penalized lasso regression method, often used in machine learning to select the subset of variables. Extra Trees Classifier [ 82 ] is an example of a tree-based estimator that can be used to compute impurity-based function importance, which can then be used to discard irrelevant features.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig8_HTML.jpg

An example of a principal component analysis (PCA) and created principal components PC1 and PC2 in different dimension space

Association Rule Learning

Association rule learning is a rule-based machine learning approach to discover interesting relationships, “IF-THEN” statements, in large datasets between variables [ 7 ]. One example is that “if a customer buys a computer or laptop (an item), s/he is likely to also buy anti-virus software (another item) at the same time”. Association rules are employed today in many application areas, including IoT services, medical diagnosis, usage behavior analytics, web usage mining, smartphone applications, cybersecurity applications, and bioinformatics. In comparison to sequence mining, association rule learning does not usually take into account the order of things within or across transactions. A common way of measuring the usefulness of association rules is to use its parameter, the ‘support’ and ‘confidence’, which is introduced in [ 7 ].

In the data mining literature, many association rule learning methods have been proposed, such as logic dependent [ 34 ], frequent pattern based [ 8 , 49 , 68 ], and tree-based [ 42 ]. The most popular association rule learning algorithms are summarized below.

  • AIS and SETM: AIS is the first algorithm proposed by Agrawal et al. [ 7 ] for association rule mining. The AIS algorithm’s main downside is that too many candidate itemsets are generated, requiring more space and wasting a lot of effort. This algorithm calls for too many passes over the entire dataset to produce the rules. Another approach SETM [ 49 ] exhibits good performance and stable behavior with execution time; however, it suffers from the same flaw as the AIS algorithm.
  • Apriori: For generating association rules for a given dataset, Agrawal et al. [ 8 ] proposed the Apriori, Apriori-TID, and Apriori-Hybrid algorithms. These later algorithms outperform the AIS and SETM mentioned above due to the Apriori property of frequent itemset [ 8 ]. The term ‘Apriori’ usually refers to having prior knowledge of frequent itemset properties. Apriori uses a “bottom-up” approach, where it generates the candidate itemsets. To reduce the search space, Apriori uses the property “all subsets of a frequent itemset must be frequent; and if an itemset is infrequent, then all its supersets must also be infrequent”. Another approach predictive Apriori [ 108 ] can also generate rules; however, it receives unexpected results as it combines both the support and confidence. The Apriori [ 8 ] is the widely applicable techniques in mining association rules.
  • ECLAT: This technique was proposed by Zaki et al. [ 131 ] and stands for Equivalence Class Clustering and bottom-up Lattice Traversal. ECLAT uses a depth-first search to find frequent itemsets. In contrast to the Apriori [ 8 ] algorithm, which represents data in a horizontal pattern, it represents data vertically. Hence, the ECLAT algorithm is more efficient and scalable in the area of association rule learning. This algorithm is better suited for small and medium datasets whereas the Apriori algorithm is used for large datasets.
  • FP-Growth: Another common association rule learning technique based on the frequent-pattern tree (FP-tree) proposed by Han et al. [ 42 ] is Frequent Pattern Growth, known as FP-Growth. The key difference with Apriori is that while generating rules, the Apriori algorithm [ 8 ] generates frequent candidate itemsets; on the other hand, the FP-growth algorithm [ 42 ] prevents candidate generation and thus produces a tree by the successful strategy of ‘divide and conquer’ approach. Due to its sophistication, however, FP-Tree is challenging to use in an interactive mining environment [ 133 ]. Thus, the FP-Tree would not fit into memory for massive data sets, making it challenging to process big data as well. Another solution is RARM (Rapid Association Rule Mining) proposed by Das et al. [ 26 ] but faces a related FP-tree issue [ 133 ].
  • ABC-RuleMiner: A rule-based machine learning method, recently proposed in our earlier paper, by Sarker et al. [ 104 ], to discover the interesting non-redundant rules to provide real-world intelligent services. This algorithm effectively identifies the redundancy in associations by taking into account the impact or precedence of the related contextual features and discovers a set of non-redundant association rules. This algorithm first constructs an association generation tree (AGT), a top-down approach, and then extracts the association rules through traversing the tree. Thus, ABC-RuleMiner is more potent than traditional rule-based methods in terms of both non-redundant rule generation and intelligent decision-making, particularly in a context-aware smart computing environment, where human or user preferences are involved.

Among the association rule learning techniques discussed above, Apriori [ 8 ] is the most widely used algorithm for discovering association rules from a given dataset [ 133 ]. The main strength of the association learning technique is its comprehensiveness, as it generates all associations that satisfy the user-specified constraints, such as minimum support and confidence value. The ABC-RuleMiner approach [ 104 ] discussed earlier could give significant results in terms of non-redundant rule generation and intelligent decision-making for the relevant application areas in the real world.

Reinforcement Learning

Reinforcement learning (RL) is a machine learning technique that allows an agent to learn by trial and error in an interactive environment using input from its actions and experiences. Unlike supervised learning, which is based on given sample data or examples, the RL method is based on interacting with the environment. The problem to be solved in reinforcement learning (RL) is defined as a Markov Decision Process (MDP) [ 86 ], i.e., all about sequentially making decisions. An RL problem typically includes four elements such as Agent, Environment, Rewards, and Policy.

RL can be split roughly into Model-based and Model-free techniques. Model-based RL is the process of inferring optimal behavior from a model of the environment by performing actions and observing the results, which include the next state and the immediate reward [ 85 ]. AlphaZero, AlphaGo [ 113 ] are examples of the model-based approaches. On the other hand, a model-free approach does not use the distribution of the transition probability and the reward function associated with MDP. Q-learning, Deep Q Network, Monte Carlo Control, SARSA (State–Action–Reward–State–Action), etc. are some examples of model-free algorithms [ 52 ]. The policy network, which is required for model-based RL but not for model-free, is the key difference between model-free and model-based learning. In the following, we discuss the popular RL algorithms.

  • Monte Carlo methods: Monte Carlo techniques, or Monte Carlo experiments, are a wide category of computational algorithms that rely on repeated random sampling to obtain numerical results [ 52 ]. The underlying concept is to use randomness to solve problems that are deterministic in principle. Optimization, numerical integration, and making drawings from the probability distribution are the three problem classes where Monte Carlo techniques are most commonly used.
  • Q-learning: Q-learning is a model-free reinforcement learning algorithm for learning the quality of behaviors that tell an agent what action to take under what conditions [ 52 ]. It does not need a model of the environment (hence the term “model-free”), and it can deal with stochastic transitions and rewards without the need for adaptations. The ‘Q’ in Q-learning usually stands for quality, as the algorithm calculates the maximum expected rewards for a given behavior in a given state.
  • Deep Q-learning: The basic working step in Deep Q-Learning [ 52 ] is that the initial state is fed into the neural network, which returns the Q-value of all possible actions as an output. Still, when we have a reasonably simple setting to overcome, Q-learning works well. However, when the number of states and actions becomes more complicated, deep learning can be used as a function approximator.

Reinforcement learning, along with supervised and unsupervised learning, is one of the basic machine learning paradigms. RL can be used to solve numerous real-world problems in various fields, such as game theory, control theory, operations analysis, information theory, simulation-based optimization, manufacturing, supply chain logistics, multi-agent systems, swarm intelligence, aircraft control, robot motion control, and many more.

Artificial Neural Network and Deep Learning

Deep learning is part of a wider family of artificial neural networks (ANN)-based machine learning approaches with representation learning. Deep learning provides a computational architecture by combining several processing layers, such as input, hidden, and output layers, to learn from data [ 41 ]. The main advantage of deep learning over traditional machine learning methods is its better performance in several cases, particularly learning from large datasets [ 105 , 129 ]. Figure ​ Figure9 9 shows a general performance of deep learning over machine learning considering the increasing amount of data. However, it may vary depending on the data characteristics and experimental set up.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig9_HTML.jpg

Machine learning and deep learning performance in general with the amount of data

The most common deep learning algorithms are: Multi-layer Perceptron (MLP), Convolutional Neural Network (CNN, or ConvNet), Long Short-Term Memory Recurrent Neural Network (LSTM-RNN) [ 96 ]. In the following, we discuss various types of deep learning methods that can be used to build effective data-driven models for various purposes.

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig10_HTML.jpg

A structure of an artificial neural network modeling with multiple processing layers

An external file that holds a picture, illustration, etc.
Object name is 42979_2021_592_Fig11_HTML.jpg

An example of a convolutional neural network (CNN or ConvNet) including multiple convolution and pooling layers

  • LSTM-RNN: Long short-term memory (LSTM) is an artificial recurrent neural network (RNN) architecture used in the area of deep learning [ 38 ]. LSTM has feedback links, unlike normal feed-forward neural networks. LSTM networks are well-suited for analyzing and learning sequential data, such as classifying, processing, and predicting data based on time series data, which differentiates it from other conventional networks. Thus, LSTM can be used when the data are in a sequential format, such as time, sentence, etc., and commonly applied in the area of time-series analysis, natural language processing, speech recognition, etc.

In addition to these most common deep learning methods discussed above, several other deep learning approaches [ 96 ] exist in the area for various purposes. For instance, the self-organizing map (SOM) [ 58 ] uses unsupervised learning to represent the high-dimensional data by a 2D grid map, thus achieving dimensionality reduction. The autoencoder (AE) [ 15 ] is another learning technique that is widely used for dimensionality reduction as well and feature extraction in unsupervised learning tasks. Restricted Boltzmann machines (RBM) [ 46 ] can be used for dimensionality reduction, classification, regression, collaborative filtering, feature learning, and topic modeling. A deep belief network (DBN) is typically composed of simple, unsupervised networks such as restricted Boltzmann machines (RBMs) or autoencoders, and a backpropagation neural network (BPNN) [ 123 ]. A generative adversarial network (GAN) [ 39 ] is a form of the network for deep learning that can generate data with characteristics close to the actual data input. Transfer learning is currently very common because it can train deep neural networks with comparatively low data, which is typically the re-use of a new problem with a pre-trained model [ 124 ]. A brief discussion of these artificial neural networks (ANN) and deep learning (DL) models are summarized in our earlier paper Sarker et al. [ 96 ].

Overall, based on the learning techniques discussed above, we can conclude that various types of machine learning techniques, such as classification analysis, regression, data clustering, feature selection and extraction, and dimensionality reduction, association rule learning, reinforcement learning, or deep learning techniques, can play a significant role for various purposes according to their capabilities. In the following section, we discuss several application areas based on machine learning algorithms.

Applications of Machine Learning

In the current age of the Fourth Industrial Revolution (4IR), machine learning becomes popular in various application areas, because of its learning capabilities from the past and making intelligent decisions. In the following, we summarize and discuss ten popular application areas of machine learning technology.

  • Predictive analytics and intelligent decision-making: A major application field of machine learning is intelligent decision-making by data-driven predictive analytics [ 21 , 70 ]. The basis of predictive analytics is capturing and exploiting relationships between explanatory variables and predicted variables from previous events to predict the unknown outcome [ 41 ]. For instance, identifying suspects or criminals after a crime has been committed, or detecting credit card fraud as it happens. Another application, where machine learning algorithms can assist retailers in better understanding consumer preferences and behavior, better manage inventory, avoiding out-of-stock situations, and optimizing logistics and warehousing in e-commerce. Various machine learning algorithms such as decision trees, support vector machines, artificial neural networks, etc. [ 106 , 125 ] are commonly used in the area. Since accurate predictions provide insight into the unknown, they can improve the decisions of industries, businesses, and almost any organization, including government agencies, e-commerce, telecommunications, banking and financial services, healthcare, sales and marketing, transportation, social networking, and many others.
  • Cybersecurity and threat intelligence: Cybersecurity is one of the most essential areas of Industry 4.0. [ 114 ], which is typically the practice of protecting networks, systems, hardware, and data from digital attacks [ 114 ]. Machine learning has become a crucial cybersecurity technology that constantly learns by analyzing data to identify patterns, better detect malware in encrypted traffic, find insider threats, predict where bad neighborhoods are online, keep people safe while browsing, or secure data in the cloud by uncovering suspicious activity. For instance, clustering techniques can be used to identify cyber-anomalies, policy violations, etc. To detect various types of cyber-attacks or intrusions machine learning classification models by taking into account the impact of security features are useful [ 97 ]. Various deep learning-based security models can also be used on the large scale of security datasets [ 96 , 129 ]. Moreover, security policy rules generated by association rule learning techniques can play a significant role to build a rule-based security system [ 105 ]. Thus, we can say that various learning techniques discussed in Sect. Machine Learning Tasks and Algorithms , can enable cybersecurity professionals to be more proactive inefficiently preventing threats and cyber-attacks.
  • Internet of things (IoT) and smart cities: Internet of Things (IoT) is another essential area of Industry 4.0. [ 114 ], which turns everyday objects into smart objects by allowing them to transmit data and automate tasks without the need for human interaction. IoT is, therefore, considered to be the big frontier that can enhance almost all activities in our lives, such as smart governance, smart home, education, communication, transportation, retail, agriculture, health care, business, and many more [ 70 ]. Smart city is one of IoT’s core fields of application, using technologies to enhance city services and residents’ living experiences [ 132 , 135 ]. As machine learning utilizes experience to recognize trends and create models that help predict future behavior and events, it has become a crucial technology for IoT applications [ 103 ]. For example, to predict traffic in smart cities, parking availability prediction, estimate the total usage of energy of the citizens for a particular period, make context-aware and timely decisions for the people, etc. are some tasks that can be solved using machine learning techniques according to the current needs of the people.
  • Traffic prediction and transportation: Transportation systems have become a crucial component of every country’s economic development. Nonetheless, several cities around the world are experiencing an excessive rise in traffic volume, resulting in serious issues such as delays, traffic congestion, higher fuel prices, increased CO 2 pollution, accidents, emergencies, and a decline in modern society’s quality of life [ 40 ]. Thus, an intelligent transportation system through predicting future traffic is important, which is an indispensable part of a smart city. Accurate traffic prediction based on machine and deep learning modeling can help to minimize the issues [ 17 , 30 , 31 ]. For example, based on the travel history and trend of traveling through various routes, machine learning can assist transportation companies in predicting possible issues that may occur on specific routes and recommending their customers to take a different path. Ultimately, these learning-based data-driven models help improve traffic flow, increase the usage and efficiency of sustainable modes of transportation, and limit real-world disruption by modeling and visualizing future changes.
  • Healthcare and COVID-19 pandemic: Machine learning can help to solve diagnostic and prognostic problems in a variety of medical domains, such as disease prediction, medical knowledge extraction, detecting regularities in data, patient management, etc. [ 33 , 77 , 112 ]. Coronavirus disease (COVID-19) is an infectious disease caused by a newly discovered coronavirus, according to the World Health Organization (WHO) [ 3 ]. Recently, the learning techniques have become popular in the battle against COVID-19 [ 61 , 63 ]. For the COVID-19 pandemic, the learning techniques are used to classify patients at high risk, their mortality rate, and other anomalies [ 61 ]. It can also be used to better understand the virus’s origin, COVID-19 outbreak prediction, as well as for disease diagnosis and treatment [ 14 , 50 ]. With the help of machine learning, researchers can forecast where and when, the COVID-19 is likely to spread, and notify those regions to match the required arrangements. Deep learning also provides exciting solutions to the problems of medical image processing and is seen as a crucial technique for potential applications, particularly for COVID-19 pandemic [ 10 , 78 , 111 ]. Overall, machine and deep learning techniques can help to fight the COVID-19 virus and the pandemic as well as intelligent clinical decisions making in the domain of healthcare.
  • E-commerce and product recommendations: Product recommendation is one of the most well known and widely used applications of machine learning, and it is one of the most prominent features of almost any e-commerce website today. Machine learning technology can assist businesses in analyzing their consumers’ purchasing histories and making customized product suggestions for their next purchase based on their behavior and preferences. E-commerce companies, for example, can easily position product suggestions and offers by analyzing browsing trends and click-through rates of specific items. Using predictive modeling based on machine learning techniques, many online retailers, such as Amazon [ 71 ], can better manage inventory, prevent out-of-stock situations, and optimize logistics and warehousing. The future of sales and marketing is the ability to capture, evaluate, and use consumer data to provide a customized shopping experience. Furthermore, machine learning techniques enable companies to create packages and content that are tailored to the needs of their customers, allowing them to maintain existing customers while attracting new ones.
  • NLP and sentiment analysis: Natural language processing (NLP) involves the reading and understanding of spoken or written language through the medium of a computer [ 79 , 103 ]. Thus, NLP helps computers, for instance, to read a text, hear speech, interpret it, analyze sentiment, and decide which aspects are significant, where machine learning techniques can be used. Virtual personal assistant, chatbot, speech recognition, document description, language or machine translation, etc. are some examples of NLP-related tasks. Sentiment Analysis [ 90 ] (also referred to as opinion mining or emotion AI) is an NLP sub-field that seeks to identify and extract public mood and views within a given text through blogs, reviews, social media, forums, news, etc. For instance, businesses and brands use sentiment analysis to understand the social sentiment of their brand, product, or service through social media platforms or the web as a whole. Overall, sentiment analysis is considered as a machine learning task that analyzes texts for polarity, such as “positive”, “negative”, or “neutral” along with more intense emotions like very happy, happy, sad, very sad, angry, have interest, or not interested etc.
  • Image, speech and pattern recognition: Image recognition [ 36 ] is a well-known and widespread example of machine learning in the real world, which can identify an object as a digital image. For instance, to label an x-ray as cancerous or not, character recognition, or face detection in an image, tagging suggestions on social media, e.g., Facebook, are common examples of image recognition. Speech recognition [ 23 ] is also very popular that typically uses sound and linguistic models, e.g., Google Assistant, Cortana, Siri, Alexa, etc. [ 67 ], where machine learning methods are used. Pattern recognition [ 13 ] is defined as the automated recognition of patterns and regularities in data, e.g., image analysis. Several machine learning techniques such as classification, feature selection, clustering, or sequence labeling methods are used in the area.
  • Sustainable agriculture: Agriculture is essential to the survival of all human activities [ 109 ]. Sustainable agriculture practices help to improve agricultural productivity while also reducing negative impacts on the environment [ 5 , 25 , 109 ]. The sustainable agriculture supply chains are knowledge-intensive and based on information, skills, technologies, etc., where knowledge transfer encourages farmers to enhance their decisions to adopt sustainable agriculture practices utilizing the increasing amount of data captured by emerging technologies, e.g., the Internet of Things (IoT), mobile technologies and devices, etc. [ 5 , 53 , 54 ]. Machine learning can be applied in various phases of sustainable agriculture, such as in the pre-production phase - for the prediction of crop yield, soil properties, irrigation requirements, etc.; in the production phase—for weather prediction, disease detection, weed detection, soil nutrient management, livestock management, etc.; in processing phase—for demand estimation, production planning, etc. and in the distribution phase - the inventory management, consumer analysis, etc.
  • User behavior analytics and context-aware smartphone applications: Context-awareness is a system’s ability to capture knowledge about its surroundings at any moment and modify behaviors accordingly [ 28 , 93 ]. Context-aware computing uses software and hardware to automatically collect and interpret data for direct responses. The mobile app development environment has been changed greatly with the power of AI, particularly, machine learning techniques through their learning capabilities from contextual data [ 103 , 136 ]. Thus, the developers of mobile apps can rely on machine learning to create smart apps that can understand human behavior, support, and entertain users [ 107 , 137 , 140 ]. To build various personalized data-driven context-aware systems, such as smart interruption management, smart mobile recommendation, context-aware smart searching, decision-making that intelligently assist end mobile phone users in a pervasive computing environment, machine learning techniques are applicable. For example, context-aware association rules can be used to build an intelligent phone call application [ 104 ]. Clustering approaches are useful in capturing users’ diverse behavioral activities by taking into account data in time series [ 102 ]. To predict the future events in various contexts, the classification methods can be used [ 106 , 139 ]. Thus, various learning techniques discussed in Sect. “ Machine Learning Tasks and Algorithms ” can help to build context-aware adaptive and smart applications according to the preferences of the mobile phone users.

In addition to these application areas, machine learning-based models can also apply to several other domains such as bioinformatics, cheminformatics, computer networks, DNA sequence classification, economics and banking, robotics, advanced engineering, and many more.

Challenges and Research Directions

Our study on machine learning algorithms for intelligent data analysis and applications opens several research issues in the area. Thus, in this section, we summarize and discuss the challenges faced and the potential research opportunities and future directions.

In general, the effectiveness and the efficiency of a machine learning-based solution depend on the nature and characteristics of the data, and the performance of the learning algorithms. To collect the data in the relevant domain, such as cybersecurity, IoT, healthcare and agriculture discussed in Sect. “ Applications of Machine Learning ” is not straightforward, although the current cyberspace enables the production of a huge amount of data with very high frequency. Thus, collecting useful data for the target machine learning-based applications, e.g., smart city applications, and their management is important to further analysis. Therefore, a more in-depth investigation of data collection methods is needed while working on the real-world data. Moreover, the historical data may contain many ambiguous values, missing values, outliers, and meaningless data. The machine learning algorithms, discussed in Sect “ Machine Learning Tasks and Algorithms ” highly impact on data quality, and availability for training, and consequently on the resultant model. Thus, to accurately clean and pre-process the diverse data collected from diverse sources is a challenging task. Therefore, effectively modifying or enhance existing pre-processing methods, or proposing new data preparation techniques are required to effectively use the learning algorithms in the associated application domain.

To analyze the data and extract insights, there exist many machine learning algorithms, summarized in Sect. “ Machine Learning Tasks and Algorithms ”. Thus, selecting a proper learning algorithm that is suitable for the target application is challenging. The reason is that the outcome of different learning algorithms may vary depending on the data characteristics [ 106 ]. Selecting a wrong learning algorithm would result in producing unexpected outcomes that may lead to loss of effort, as well as the model’s effectiveness and accuracy. In terms of model building, the techniques discussed in Sect. “ Machine Learning Tasks and Algorithms ” can directly be used to solve many real-world issues in diverse domains, such as cybersecurity, smart cities and healthcare summarized in Sect. “ Applications of Machine Learning ”. However, the hybrid learning model, e.g., the ensemble of methods, modifying or enhancement of the existing learning techniques, or designing new learning methods, could be a potential future work in the area.

Thus, the ultimate success of a machine learning-based solution and corresponding applications mainly depends on both the data and the learning algorithms. If the data are bad to learn, such as non-representative, poor-quality, irrelevant features, or insufficient quantity for training, then the machine learning models may become useless or will produce lower accuracy. Therefore, effectively processing the data and handling the diverse learning algorithms are important, for a machine learning-based solution and eventually building intelligent applications.

In this paper, we have conducted a comprehensive overview of machine learning algorithms for intelligent data analysis and applications. According to our goal, we have briefly discussed how various types of machine learning methods can be used for making solutions to various real-world issues. A successful machine learning model depends on both the data and the performance of the learning algorithms. The sophisticated learning algorithms then need to be trained through the collected real-world data and knowledge related to the target application before the system can assist with intelligent decision-making. We also discussed several popular application areas based on machine learning techniques to highlight their applicability in various real-world issues. Finally, we have summarized and discussed the challenges faced and the potential research opportunities and future directions in the area. Therefore, the challenges that are identified create promising research opportunities in the field which must be addressed with effective solutions in various application areas. Overall, we believe that our study on machine learning-based solutions opens up a promising direction and can be used as a reference guide for potential research and applications for both academia and industry professionals as well as for decision-makers, from a technical point of view.


The author declares no conflict of interest.

This article is part of the topical collection “Advances in Computational Approaches for Artificial Intelligence, Image Processing, IoT and Cloud Applications” guest edited by Bhanu Prakash K N and M. Shivakumar.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

research paper machine learning

Machine Learning

  • An international forum for research on computational approaches to learning.
  • Reports substantive results on a wide range of learning methods applied to a variety of learning problems.
  • Provides solid support via empirical studies, theoretical analysis, or comparison to psychological phenomena.
  • Shows how to apply learning methods to solve important applications problems.
  • Improves how machine learning research is conducted.

Journal information

  • Hendrik Blockeel

Journal metrics

Latest issue.

research paper machine learning

Issue 12, December 2023

Latest articles

Predicting potential real-time donations in youtube live streaming services via continuous-time dynamic graphs.

  • Ruidong Jin
  • Tsuyoshi Murata
  • Content type: OriginalPaper
  • Open Access
  • Published: 10 November 2023
  • Special Issue on Discovery Science 2022

research paper machine learning

Temporal silhouette: validation of stream clustering robust to concept drift

  • Félix Iglesias Vázquez
  • Tanja Zseby

research paper machine learning

Optimizing model-agnostic random subspace ensembles

  • Vân Anh Huynh-Thu
  • Pierre Geurts
  • Published: 08 November 2023
  • Special Issue of the ECML PKDD 2023 Journal Track

research paper machine learning

GVFs in the real world: making predictions online for water treatment

Authors (first, second and last of 6).

  • Muhammad Kamran Janjua
  • Haseeb Shah
  • Reinforcement Learning for Real Life

research paper machine learning

Efficient learning of power grid voltage control strategies via model-based deep reinforcement learning

Authors (first, second and last of 8).

  • Ramij Raja Hossain
  • Tianzhixi Yin
  • Qiuhua Huang
  • Published: 06 November 2023
  • Special Issue on Reinforcement Learning for Real Life

research paper machine learning

Journal updates

Cfp: ijclr learning and reasoning, call for papers: conformal prediction and distribution-free uncertainty quantification.

Submission Deadline: November 30, 2023 

Guest Editors: Henrik Boström, Eyke Hüllermeier, Ulf Johansson, Khuong An Nguyen, Aaditya Ramdas

CfP: Special Issue on ACML 2023

Guest editors: Vu Nguyen, Dani Yogatama

Submission deadline: June 2, 2023

CfP: Emerging Applications and Frontiers for Data Science (DSAA 2023 journal track)

Submission Deadline: April 15th, 2023

Guest Editors: Bingxue Zhang (Lead GE), Feida Zhu, Bin Yang, João Gama

For authors

Working on a manuscript.

Avoid the most common mistakes and prepare your manuscript for journal editors.

About this journal

  • ACM Digital Library
  • Current Contents/Engineering, Computing and Technology
  • EBSCO Applied Science & Technology Source
  • EBSCO Associates Programs Source
  • EBSCO Book Review Digest Plus
  • EBSCO Computer Science Index
  • EBSCO Computers & Applied Sciences Complete
  • EBSCO Discovery Service
  • EBSCO Engineering Source
  • EBSCO Linguistics Abstracts Online
  • EBSCO Military Transition Support Center
  • EBSCO OmniFile
  • EBSCO STM Source
  • EBSCO Science Full Text Select
  • EBSCO Vocational Studies
  • EI Compendex
  • Google Scholar
  • Japanese Science and Technology Agency (JST)
  • Journal Citation Reports/Science Edition
  • Mathematical Reviews
  • OCLC WorldCat Discovery Service
  • ProQuest Advanced Technologies & Aerospace Database
  • ProQuest-ExLibris Primo
  • ProQuest-ExLibris Summon
  • Science Citation Index
  • Science Citation Index Expanded (SCIE)
  • TD Net Discovery Service
  • UGC-CARE List (India)

Rights and permissions

Springer policies

© Springer Science+Business Media LLC, part of Springer Nature

research paper machine learning

Frequently Asked Questions

Journal of Machine Learning Research

The Journal of Machine Learning Research (JMLR), established in 2000 , provides an international forum for the electronic and paper publication of high-quality scholarly articles in all areas of machine learning. All published papers are freely available online.

  • 2023.01.20 : Volume 23 completed; Volume 24 began.
  • 2022.07.20 : New special issue on climate change .
  • 2022.02.18 : New blog post: Retrospectives from 20 Years of JMLR .
  • 2022.01.25 : Volume 22 completed; Volume 23 began.
  • 2021.12.02 : Message from outgoing co-EiC Bernhard Schölkopf .
  • 2021.02.10 : Volume 21 completed; Volume 22 began.
  • More news ...

Latest papers

Bagging in overparameterized learning: Risk characterization and risk monotonization Pratik Patil, Jin-Hong Du, Arun Kumar Kuchibhotla , 2023. [ abs ][ pdf ][ bib ]

Operator learning with PCA-Net: upper and lower complexity bounds Samuel Lanthaler , 2023. [ abs ][ pdf ][ bib ]

Mixed Regression via Approximate Message Passing Nelvin Tan, Ramji Venkataramanan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

The Dynamics of Sharpness-Aware Minimization: Bouncing Across Ravines and Drifting Towards Wide Minima Peter L. Bartlett, Philip M. Long, Olivier Bousquet , 2023. [ abs ][ pdf ][ bib ]

MARLlib: A Scalable and Efficient Multi-agent Reinforcement Learning Library Siyi Hu, Yifan Zhong, Minquan Gao, Weixun Wang, Hao Dong, Xiaodan Liang, Zhihui Li, Xiaojun Chang, Yaodong Yang , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Fast Expectation Propagation for Heteroscedastic, Lasso-Penalized, and Quantile Regression Jackson Zhou, John T. Ormerod, Clara Grazian , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Zeroth-Order Alternating Gradient Descent Ascent Algorithms for A Class of Nonconvex-Nonconcave Minimax Problems Zi Xu, Zi-Qi Wang, Jun-Lin Wang, Yu-Hong Dai , 2023. [ abs ][ pdf ][ bib ]

The Measure and Mismeasure of Fairness Sam Corbett-Davies, Johann D. Gaebler, Hamed Nilforoshan, Ravi Shroff, Sharad Goel , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Microcanonical Hamiltonian Monte Carlo Jakob Robnik, G. Bruno De Luca, Eva Silverstein, Uroš Seljak , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Prediction Equilibrium for Dynamic Network Flows Lukas Graf, Tobias Harks, Kostas Kollias, Michael Markl , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Dimension Reduction and MARS Yu Liu LIU, Degui Li, Yingcun Xia , 2023. [ abs ][ pdf ][ bib ]

Nevis'22: A Stream of 100 Tasks Sampled from 30 Years of Computer Vision Research Jorg Bornschein, Alexandre Galashov, Ross Hemsley, Amal Rannen-Triki, Yutian Chen, Arslan Chaudhry, Xu Owen He, Arthur Douillard, Massimo Caccia, Qixuan Feng, Jiajun Shen, Sylvestre-Alvise Rebuffi, Kitty Stacpoole, Diego de las Casas, Will Hawkins, Angeliki Lazaridou, Yee Whye Teh, Andrei A. Rusu, Razvan Pascanu, Marc’Aurelio Ranzato , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Fast Screening Rules for Optimal Design via Quadratic Lasso Reformulation Guillaume Sagnol, Luc Pronzato , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Multi-Consensus Decentralized Accelerated Gradient Descent Haishan Ye, Luo Luo, Ziang Zhou, Tong Zhang , 2023. [ abs ][ pdf ][ bib ]

Continuous-in-time Limit for Bayesian Bandits Yuhua Zhu, Zachary Izzo, Lexing Ying , 2023. [ abs ][ pdf ][ bib ]

Two Sample Testing in High Dimension via Maximum Mean Discrepancy Hanjia Gao, Xiaofeng Shao , 2023. [ abs ][ pdf ][ bib ]

Random Feature Amplification: Feature Learning and Generalization in Neural Networks Spencer Frei, Niladri S. Chatterji, Peter L. Bartlett , 2023. [ abs ][ pdf ][ bib ]

Pivotal Estimation of Linear Discriminant Analysis in High Dimensions Ethan X. Fang, Yajun Mei, Yuyang Shi, Qunzhi Xu, Tuo Zhao , 2023. [ abs ][ pdf ][ bib ]

Learning Optimal Feedback Operators and their Sparse Polynomial Approximations Karl Kunisch, Donato Vásquez-Varas, Daniel Walter , 2023. [ abs ][ pdf ][ bib ]

Sensitivity-Free Gradient Descent Algorithms Ion Matei, Maksym Zhenirovskyy, Johan de Kleer, John Maxwell , 2023. [ abs ][ pdf ][ bib ]

A PDE approach for regret bounds under partial monitoring Erhan Bayraktar, Ibrahim Ekren, Xin Zhang , 2023. [ abs ][ pdf ][ bib ]

A General Learning Framework for Open Ad Hoc Teamwork Using Graph-based Policy Learning Arrasy Rahman, Ignacio Carlucho, Niklas Höpner, Stefano V. Albrecht , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Causal Bandits for Linear Structural Equation Models Burak Varici, Karthikeyan Shanmugam, Prasanna Sattigeri, Ali Tajer , 2023. [ abs ][ pdf ][ bib ]

High-Dimensional Inference for Generalized Linear Models with Hidden Confounding Jing Ouyang, Kean Ming Tan, Gongjun Xu , 2023. [ abs ][ pdf ][ bib ]

Weibull Racing Survival Analysis with Competing Events, Left Truncation, and Time-Varying Covariates Quan Zhang, Yanxun Xu, Mei-Cheng Wang, Mingyuan Zhou , 2023. [ abs ][ pdf ][ bib ]

Erratum: Risk Bounds for the Majority Vote: From a PAC-Bayesian Analysis to a Learning Algorithm Louis-Philippe Vignault, Audrey Durand, Pascal Germain , 2023. [ abs ][ pdf ][ bib ]

Augmented Transfer Regression Learning with Semi-non-parametric Nuisance Models Molei Liu, Yi Zhang, Katherine P. Liao, Tianxi Cai , 2023. [ abs ][ pdf ][ bib ]

From Understanding Genetic Drift to a Smart-Restart Mechanism for Estimation-of-Distribution Algorithms Weijie Zheng, Benjamin Doerr , 2023. [ abs ][ pdf ][ bib ]

A Unified Analysis of Multi-task Functional Linear Regression Models with Manifold Constraint and Composite Quadratic Penalty Shiyuan He, Hanxuan Ye, Kejun He , 2023. [ abs ][ pdf ][ bib ]

Deletion and Insertion Tests in Regression Models Naofumi Hama, Masayoshi Mase, Art B. Owen , 2023. [ abs ][ pdf ][ bib ]

Deep Neural Networks with Dependent Weights: Gaussian Process Mixture Limit, Heavy Tails, Sparsity and Compressibility Hoil Lee, Fadhel Ayed, Paul Jung, Juho Lee, Hongseok Yang, Francois Caron , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A New Look at Dynamic Regret for Non-Stationary Stochastic Bandits Yasin Abbasi-Yadkori, Andras Gyorgy, Nevena Lazic , 2023. [ abs ][ pdf ][ bib ]

Universal Approximation Property of Invertible Neural Networks Isao Ishikawa, Takeshi Teshima, Koichi Tojo, Kenta Oono, Masahiro Ikeda, Masashi Sugiyama , 2023. [ abs ][ pdf ][ bib ]

Low Tree-Rank Bayesian Vector Autoregression Models Leo L Duan, Zeyu Yuwen, George Michailidis, Zhengwu Zhang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Generic Unsupervised Optimization for a Latent Variable Model With Exponential Family Observables Hamid Mousavi, Jakob Drefs, Florian Hirschberger, Jörg Lücke , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A Complete Characterization of Linear Estimators for Offline Policy Evaluation Juan C. Perdomo, Akshay Krishnamurthy, Peter Bartlett, Sham Kakade , 2023. [ abs ][ pdf ][ bib ]

Near-Optimal Weighted Matrix Completion Oscar López , 2023. [ abs ][ pdf ][ bib ]

Community models for networks observed through edge nominations Tianxi Li, Elizaveta Levina, Ji Zhu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

The Bayesian Learning Rule Mohammad Emtiyaz Khan, Håvard Rue , 2023. [ abs ][ pdf ][ bib ]

Removing Data Heterogeneity Influence Enhances Network Topology Dependence of Decentralized SGD Kun Yuan, Sulaiman A. Alghunaim, Xinmeng Huang , 2023. [ abs ][ pdf ][ bib ]

Sparse Markov Models for High-dimensional Inference Guilherme Ost, Daniel Y. Takahashi , 2023. [ abs ][ pdf ][ bib ]

Distinguishing Cause and Effect in Bivariate Structural Causal Models: A Systematic Investigation Christoph Käding,, Jakob Runge, , 2023. [ abs ][ pdf ][ bib ]

Elastic Gradient Descent, an Iterative Optimization Method Approximating the Solution Paths of the Elastic Net Oskar Allerbo, Johan Jonasson, Rebecka Jörnsten , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On Biased Compression for Distributed Learning Aleksandr Beznosikov, Samuel Horváth, Peter Richtárik, Mher Safaryan , 2023. [ abs ][ pdf ][ bib ]

Adaptive Clustering Using Kernel Density Estimators Ingo Steinwart, Bharath K. Sriperumbudur, Philipp Thomann , 2023. [ abs ][ pdf ][ bib ]

A Continuous-time Stochastic Gradient Descent Method for Continuous Data Kexin Jin, Jonas Latz, Chenguang Liu, Carola-Bibiane Schönlieb , 2023. [ abs ][ pdf ][ bib ]

Online Non-stochastic Control with Partial Feedback Yu-Hu Yan, Peng Zhao, Zhi-Hua Zhou , 2023. [ abs ][ pdf ][ bib ]

Distributed Sparse Regression via Penalization Yao Ji, Gesualdo Scutari, Ying Sun, Harsha Honnappa , 2023. [ abs ][ pdf ][ bib ]

Causal Discovery with Unobserved Confounding and Non-Gaussian Data Y. Samuel Wang, Mathias Drton , 2023. [ abs ][ pdf ][ bib ]

Sharper Analysis for Minibatch Stochastic Proximal Point Methods: Stability, Smoothness, and Deviation Xiao-Tong Yuan, Ping Li , 2023. [ abs ][ pdf ][ bib ]

Dynamic Ranking with the BTL Model: A Nearest Neighbor based Rank Centrality Method Eglantine Karlé, Hemant Tyagi , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Revisiting minimum description length complexity in overparameterized models Raaz Dwivedi, Chandan Singh, Bin Yu, Martin Wainwright , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Sparse Plus Low Rank Matrix Decomposition: A Discrete Optimization Approach Dimitris Bertsimas, Ryan Cory-Wright, Nicholas A. G. Johnson , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On the Estimation of Derivatives Using Plug-in Kernel Ridge Regression Estimators Zejian Liu, Meng Li , 2023. [ abs ][ pdf ][ bib ]

Surrogate Assisted Semi-supervised Inference for High Dimensional Risk Prediction Jue Hou, Zijian Guo, Tianxi Cai , 2023. [ abs ][ pdf ][ bib ]

ProtoryNet - Interpretable Text Classification Via Prototype Trajectories Dat Hong, Tong Wang, Stephen Baek , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Distributed Algorithms for U-statistics-based Empirical Risk Minimization Lanjue Chen, Alan T.K. Wan, Shuyi Zhang, Yong Zhou , 2023. [ abs ][ pdf ][ bib ]

Minimax Estimation for Personalized Federated Learning: An Alternative between FedAvg and Local Training? Shuxiao Chen, Qinqing Zheng, Qi Long, Weijie J. Su , 2023. [ abs ][ pdf ][ bib ]

Nearest Neighbor Dirichlet Mixtures Shounak Chattopadhyay, Antik Chakraborty, David B. Dunson , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Learning to Rank under Multinomial Logit Choice James A. Grant, David S. Leslie , 2023. [ abs ][ pdf ][ bib ]

Scalable high-dimensional Bayesian varying coefficient models with unknown within-subject covariance Ray Bai, Mary R. Boland, Yong Chen , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Multi-view Collaborative Gaussian Process Dynamical Systems Shiliang Sun, Jingjing Fei, Jing Zhao, Liang Mao , 2023. [ abs ][ pdf ][ bib ]

Fairlearn: Assessing and Improving Fairness of AI Systems Hilde Weerts, Miroslav Dudík, Richard Edgar, Adrin Jalali, Roman Lutz, Michael Madaio , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Scalable Real-Time Recurrent Learning Using Columnar-Constructive Networks Khurram Javed, Haseeb Shah, Richard S. Sutton, Martha White , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Torchhd: An Open Source Python Library to Support Research on Hyperdimensional Computing and Vector Symbolic Architectures Mike Heddes, Igor Nunes, Pere Vergés, Denis Kleyko, Danny Abraham, Tony Givargis, Alexandru Nicolau, Alexander Veidenbaum , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

skrl: Modular and Flexible Library for Reinforcement Learning Antonio Serrano-Muñoz, Dimitrios Chrysostomou, Simon Bøgh, Nestor Arana-Arexolaleiba , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model Alexandra Sasha Luccioni, Sylvain Viguier, Anne-Laure Ligozat , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Adaptive False Discovery Rate Control with Privacy Guarantee Xintao Xia, Zhanrui Cai , 2023. [ abs ][ pdf ][ bib ]

Atlas: Few-shot Learning with Retrieval Augmented Language Models Gautier Izacard, Patrick Lewis, Maria Lomeli, Lucas Hosseini, Fabio Petroni, Timo Schick, Jane Dwivedi-Yu, Armand Joulin, Sebastian Riedel, Edouard Grave , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Convex Reinforcement Learning in Finite Trials Mirco Mutti, Riccardo De Santi, Piersilvio De Bartolomeis, Marcello Restelli , 2023. [ abs ][ pdf ][ bib ]

Unbiased Multilevel Monte Carlo Methods for Intractable Distributions: MLMC Meets MCMC Tianze Wang, Guanyang Wang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Improving multiple-try Metropolis with local balancing Philippe Gagnon, Florian Maire, Giacomo Zanella , 2023. [ abs ][ pdf ][ bib ]

Importance Sparsification for Sinkhorn Algorithm Mengyu Li, Jun Yu, Tao Li, Cheng Meng , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Graph Attention Retrospective Kimon Fountoulakis, Amit Levi, Shenghao Yang, Aseem Baranwal, Aukosh Jagannath , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Confidence Intervals and Hypothesis Testing for High-dimensional Quantile Regression: Convolution Smoothing and Debiasing Yibo Yan, Xiaozhou Wang, Riquan Zhang , 2023. [ abs ][ pdf ][ bib ]

Selection by Prediction with Conformal p-values Ying Jin, Emmanuel J. Candes , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Alpha-divergence Variational Inference Meets Importance Weighted Auto-Encoders: Methodology and Asymptotics Kamélia Daudel, Joe Benton, Yuyang Shi, Arnaud Doucet , 2023. [ abs ][ pdf ][ bib ]

Sparse Graph Learning from Spatiotemporal Time Series Andrea Cini, Daniele Zambon, Cesare Alippi , 2023. [ abs ][ pdf ][ bib ]

Improved Powered Stochastic Optimization Algorithms for Large-Scale Machine Learning Zhuang Yang , 2023. [ abs ][ pdf ][ bib ]

PaLM: Scaling Language Modeling with Pathways Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Sebastian Gehrmann, Parker Schuh, Kensen Shi, Sasha Tsvyashchenko, Joshua Maynez, Abhishek Rao, Parker Barnes, Yi Tay, Noam Shazeer, Vinodkumar Prabhakaran, Emily Reif, Nan Du, Ben Hutchinson, Reiner Pope, James Bradbury, Jacob Austin, Michael Isard, Guy Gur-Ari, Pengcheng Yin, Toju Duke, Anselm Levskaya, Sanjay Ghemawat, Sunipa Dev, Henryk Michalewski, Xavier Garcia, Vedant Misra, Kevin Robinson, Liam Fedus, Denny Zhou, Daphne Ippolito, David Luan, Hyeontaek Lim, Barret Zoph, Alexander Spiridonov, Ryan Sepassi, David Dohan, Shivani Agrawal, Mark Omernick, Andrew M. Dai, Thanumalayan Sankaranarayana Pillai, Marie Pellat, Aitor Lewkowycz, Erica Moreira, Rewon Child, Oleksandr Polozov, Katherine Lee, Zongwei Zhou, Xuezhi Wang, Brennan Saeta, Mark Diaz, Orhan Firat, Michele Catasta, Jason Wei, Kathy Meier-Hellstern, Douglas Eck, Jeff Dean, Slav Petrov, Noah Fiedel , 2023. [ abs ][ pdf ][ bib ]

Leaky Hockey Stick Loss: The First Negatively Divergent Margin-based Loss Function for Classification Oh-Ran Kwon, Hui Zou , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Efficient Computation of Rankings from Pairwise Comparisons M. E. J. Newman , 2023. [ abs ][ pdf ][ bib ]

Scalable Computation of Causal Bounds Madhumitha Shridharan, Garud Iyengar , 2023. [ abs ][ pdf ][ bib ]

Neural Q-learning for solving PDEs Samuel N. Cohen, Deqing Jiang, Justin Sirignano , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Tractable and Near-Optimal Adversarial Algorithms for Robust Estimation in Contaminated Gaussian Models Ziyue Wang, Zhiqiang Tan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

MultiZoo and MultiBench: A Standardized Toolkit for Multimodal Deep Learning Paul Pu Liang, Yiwei Lyu, Xiang Fan, Arav Agarwal, Yun Cheng, Louis-Philippe Morency, Ruslan Salakhutdinov , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Strategic Knowledge Transfer Max Olan Smith, Thomas Anthony, Michael P. Wellman , 2023. [ abs ][ pdf ][ bib ]

Lifted Bregman Training of Neural Networks Xiaoyu Wang, Martin Benning , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Statistical Comparisons of Classifiers by Generalized Stochastic Dominance Christoph Jansen, Malte Nalenz, Georg Schollmeyer, Thomas Augustin , 2023. [ abs ][ pdf ][ bib ]

Sample Complexity for Distributionally Robust Learning under chi-square divergence Zhengyu Zhou, Weiwei Liu , 2023. [ abs ][ pdf ][ bib ]

Interpretable and Fair Boolean Rule Sets via Column Generation Connor Lawless, Sanjeeb Dash, Oktay Gunluk, Dennis Wei , 2023. [ abs ][ pdf ][ bib ]

On the Optimality of Nuclear-norm-based Matrix Completion for Problems with Smooth Non-linear Structure Yunhua Xiang, Tianyu Zhang, Xu Wang, Ali Shojaie, Noah Simon , 2023. [ abs ][ pdf ][ bib ]

Autoregressive Networks Binyan Jiang, Jialiang Li, Qiwei Yao , 2023. [ abs ][ pdf ][ bib ]

Merlion: End-to-End Machine Learning for Time Series Aadyot Bhatnagar, Paul Kassianik, Chenghao Liu, Tian Lan, Wenzhuo Yang, Rowan Cassius, Doyen Sahoo, Devansh Arpit, Sri Subramanian, Gerald Woo, Amrita Saha, Arun Kumar Jagota, Gokulakrishnan Gopalakrishnan, Manpreet Singh, K C Krithika, Sukumar Maddineni, Daeki Cho, Bo Zong, Yingbo Zhou, Caiming Xiong, Silvio Savarese, Steven Hoi, Huan Wang , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Limits of Dense Simplicial Complexes T. Mitchell Roddenberry, Santiago Segarra , 2023. [ abs ][ pdf ][ bib ]

RankSEG: A Consistent Ranking-based Framework for Segmentation Ben Dai, Chunlin Li , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Conditional Distribution Function Estimation Using Neural Networks for Censored and Uncensored Data Bingqing Hu, Bin Nan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Single Timescale Actor-Critic Method to Solve the Linear Quadratic Regulator with Convergence Guarantees Mo Zhou, Jianfeng Lu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Multi-source Learning via Completion of Block-wise Overlapping Noisy Matrices Doudou Zhou, Tianxi Cai, Junwei Lu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A Unified Framework for Factorizing Distributional Value Functions for Multi-Agent Reinforcement Learning Wei-Fang Sun, Cheng-Kuang Lee, Simon See, Chun-Yi Lee , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Functional L-Optimality Subsampling for Functional Generalized Linear Models with Massive Data Hua Liu, Jinhong You, Jiguo Cao , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Adaptation Augmented Model-based Policy Optimization Jian Shen, Hang Lai, Minghuan Liu, Han Zhao, Yong Yu, Weinan Zhang , 2023. [ abs ][ pdf ][ bib ]

GANs as Gradient Flows that Converge Yu-Jui Huang, Yuchong Zhang , 2023. [ abs ][ pdf ][ bib ]

Random Forests for Change Point Detection Malte Londschien, Peter Bühlmann, Solt Kovács , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Least Squares Model Averaging for Distributed Data Haili Zhang, Zhaobo Liu, Guohua Zou , 2023. [ abs ][ pdf ][ bib ]

An Empirical Investigation of the Role of Pre-training in Lifelong Learning Sanket Vaibhav Mehta, Darshan Patil, Sarath Chandar, Emma Strubell , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Polynomial-Time Algorithms for Counting and Sampling Markov Equivalent DAGs with Applications Marcel Wienöbst, Max Bannach, Maciej Liśkiewicz , 2023. [ abs ][ pdf ][ bib ]      [ code ]

An Inexact Augmented Lagrangian Algorithm for Training Leaky ReLU Neural Network with Group Sparsity Wei Liu, Xin Liu, Xiaojun Chen , 2023. [ abs ][ pdf ][ bib ]

Entropic Fictitious Play for Mean Field Optimization Problem Fan Chen, Zhenjie Ren, Songbo Wang , 2023. [ abs ][ pdf ][ bib ]

GFlowNet Foundations Yoshua Bengio, Salem Lahlou, Tristan Deleu, Edward J. Hu, Mo Tiwari, Emmanuel Bengio , 2023. [ abs ][ pdf ][ bib ]

LibMTL: A Python Library for Deep Multi-Task Learning Baijiong Lin, Yu Zhang , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Minimax Risk Classifiers with 0-1 Loss Santiago Mazuelas, Mauricio Romero, Peter Grunwald , 2023. [ abs ][ pdf ][ bib ]

Augmented Sparsifiers for Generalized Hypergraph Cuts Nate Veldt, Austin R. Benson, Jon Kleinberg , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Non-stationary Online Learning with Memory and Non-stochastic Control Peng Zhao, Yu-Hu Yan, Yu-Xiang Wang, Zhi-Hua Zhou , 2023. [ abs ][ pdf ][ bib ]

L0Learn: A Scalable Package for Sparse Learning using L0 Regularization Hussein Hazimeh, Rahul Mazumder, Tim Nonet , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Buffered Asynchronous SGD for Byzantine Learning Yi-Rui Yang, Wu-Jun Li , 2023. [ abs ][ pdf ][ bib ]

A Non-parametric View of FedAvg and FedProx:Beyond Stationary Points Lili Su, Jiaming Xu, Pengkun Yang , 2023. [ abs ][ pdf ][ bib ]

Multiplayer Performative Prediction: Learning in Decision-Dependent Games Adhyyan Narang, Evan Faulkner, Dmitriy Drusvyatskiy, Maryam Fazel, Lillian J. Ratliff , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Variational Inverting Network for Statistical Inverse Problems of Partial Differential Equations Junxiong Jia, Yanni Wu, Peijun Li, Deyu Meng , 2023. [ abs ][ pdf ][ bib ]

Model-based Causal Discovery for Zero-Inflated Count Data Junsouk Choi, Yang Ni , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Q-Learning for MDPs with General Spaces: Convergence and Near Optimality via Quantization under Weak Continuity Ali Kara, Naci Saldi, Serdar Yüksel , 2023. [ abs ][ pdf ][ bib ]

CodaLab Competitions: An Open Source Platform to Organize Scientific Challenges Adrien Pavao, Isabelle Guyon, Anne-Catherine Letournel, Dinh-Tuan Tran, Xavier Baro, Hugo Jair Escalante, Sergio Escalera, Tyler Thomas, Zhen Xu , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Contrasting Identifying Assumptions of Average Causal Effects: Robustness and Semiparametric Efficiency Tetiana Gorbach, Xavier de Luna, Juha Karvanen, Ingeborg Waernbaum , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Variational Gibbs Inference for Statistical Model Estimation from Incomplete Data Vaidotas Simkus, Benjamin Rhodes, Michael U. Gutmann , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Clustering and Structural Robustness in Causal Diagrams Santtu Tikka, Jouni Helske, Juha Karvanen , 2023. [ abs ][ pdf ][ bib ]      [ code ]

MMD Aggregated Two-Sample Test Antonin Schrab, Ilmun Kim, Mélisande Albert, Béatrice Laurent, Benjamin Guedj, Arthur Gretton , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Divide-and-Conquer Fusion Ryan S.Y. Chan, Murray Pollock, Adam M. Johansen, Gareth O. Roberts , 2023. [ abs ][ pdf ][ bib ]

PAC-learning for Strategic Classification Ravi Sundaram, Anil Vullikanti, Haifeng Xu, Fan Yao , 2023. [ abs ][ pdf ][ bib ]

Insights into Ordinal Embedding Algorithms: A Systematic Evaluation Leena Chennuru Vankadara, Michael Lohaus, Siavash Haghiri, Faiz Ul Wahab, Ulrike von Luxburg , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Clustering with Tangles: Algorithmic Framework and Theoretical Guarantees Solveig Klepper, Christian Elbracht, Diego Fioravanti, Jakob Kneip, Luca Rendsburg, Maximilian Teegen, Ulrike von Luxburg , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Random Feature Neural Networks Learn Black-Scholes Type PDEs Without Curse of Dimensionality Lukas Gonon , 2023. [ abs ][ pdf ][ bib ]

The Proximal ID Algorithm Ilya Shpitser, Zach Wood-Doughty, Eric J. Tchetgen Tchetgen , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Quantifying Network Similarity using Graph Cumulants Gecia Bravo-Hermsdorff, Lee M. Gunderson, Pierre-André Maugis, Carey E. Priebe , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Learning an Explicit Hyper-parameter Prediction Function Conditioned on Tasks Jun Shu, Deyu Meng, Zongben Xu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On the Theoretical Equivalence of Several Trade-Off Curves Assessing Statistical Proximity Rodrigue Siry, Ryan Webster, Loic Simon, Julien Rabin , 2023. [ abs ][ pdf ][ bib ]

Metrizing Weak Convergence with Maximum Mean Discrepancies Carl-Johann Simon-Gabriel, Alessandro Barp, Bernhard Schölkopf, Lester Mackey , 2023. [ abs ][ pdf ][ bib ]

Quasi-Equivalence between Width and Depth of Neural Networks Fenglei Fan, Rongjie Lai, Ge Wang , 2023. [ abs ][ pdf ][ bib ]

Naive regression requires weaker assumptions than factor models to adjust for multiple cause confounding Justin Grimmer, Dean Knox, Brandon Stewart , 2023. [ abs ][ pdf ][ bib ]

Factor Graph Neural Networks Zhen Zhang, Mohammed Haroon Dupty, Fan Wu, Javen Qinfeng Shi, Wee Sun Lee , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Dropout Training is Distributionally Robust Optimal José Blanchet, Yang Kang, José Luis Montiel Olea, Viet Anh Nguyen, Xuhui Zhang , 2023. [ abs ][ pdf ][ bib ]

Variational Inference for Deblending Crowded Starfields Runjing Liu, Jon D. McAuliffe, Jeffrey Regier, The LSST Dark Energy Science Collaboration , 2023. [ abs ][ pdf ][ bib ]      [ code ]

F2A2: Flexible Fully-decentralized Approximate Actor-critic for Cooperative Multi-agent Reinforcement Learning Wenhao Li, Bo Jin, Xiangfeng Wang, Junchi Yan, Hongyuan Zha , 2023. [ abs ][ pdf ][ bib ]

Comprehensive Algorithm Portfolio Evaluation using Item Response Theory Sevvandi Kandanaarachchi, Kate Smith-Miles , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Evaluating Instrument Validity using the Principle of Independent Mechanisms Patrick F. Burauel , 2023. [ abs ][ pdf ][ bib ]

Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity Kaiqing Zhang, Sham M. Kakade, Tamer Basar, Lin F. Yang , 2023. [ abs ][ pdf ][ bib ]

Posterior Consistency for Bayesian Relevance Vector Machines Xiao Fang, Malay Ghosh , 2023. [ abs ][ pdf ][ bib ]

From Classification Accuracy to Proper Scoring Rules: Elicitability of Probabilistic Top List Predictions Johannes Resin , 2023. [ abs ][ pdf ][ bib ]

Beyond the Golden Ratio for Variational Inequality Algorithms Ahmet Alacaoglu, Axel Böhm, Yura Malitsky , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Incremental Learning in Diagonal Linear Networks Raphaël Berthier , 2023. [ abs ][ pdf ][ bib ]

Small Transformers Compute Universal Metric Embeddings Anastasis Kratsios, Valentin Debarnot, Ivan Dokmanić , 2023. [ abs ][ pdf ][ bib ]      [ code ]

DART: Distance Assisted Recursive Testing Xuechan Li, Anthony D. Sung, Jichun Xie , 2023. [ abs ][ pdf ][ bib ]

Inference on the Change Point under a High Dimensional Covariance Shift Abhishek Kaul, Hongjin Zhang, Konstantinos Tsampourakis, George Michailidis , 2023. [ abs ][ pdf ][ bib ]

Bilevel Optimization with a Lower-level Contraction: Optimal Sample Complexity without Warm-Start Riccardo Grazzi, Massimiliano Pontil, Saverio Salzo , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A Parameter-Free Conditional Gradient Method for Composite Minimization under Hölder Condition Masaru Ito, Zhaosong Lu, Chuan He , 2023. [ abs ][ pdf ][ bib ]

Robust Methods for High-Dimensional Linear Learning Ibrahim Merad, Stéphane Gaïffas , 2023. [ abs ][ pdf ][ bib ]

A Framework and Benchmark for Deep Batch Active Learning for Regression David Holzmüller, Viktor Zaverkin, Johannes Kästner, Ingo Steinwart , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Preconditioned Gradient Descent for Overparameterized Nonconvex Burer--Monteiro Factorization with Global Optimality Certification Gavin Zhang, Salar Fattahi, Richard Y. Zhang , 2023. [ abs ][ pdf ][ bib ]

Flexible Model Aggregation for Quantile Regression Rasool Fakoor, Taesup Kim, Jonas Mueller, Alexander J. Smola, Ryan J. Tibshirani , 2023. [ abs ][ pdf ][ bib ]      [ code ]

q-Learning in Continuous Time Yanwei Jia, Xun Yu Zhou , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Multivariate Soft Rank via Entropy-Regularized Optimal Transport: Sample Efficiency and Generative Modeling Shoaib Bin Masud, Matthew Werenski, James M. Murphy, Shuchin Aeron , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Infinite-dimensional optimization and Bayesian nonparametric learning of stochastic differential equations Arnab Ganguly, Riten Mitra, Jinpu Zhou , 2023. [ abs ][ pdf ][ bib ]

Asynchronous Iterations in Optimization: New Sequence Results and Sharper Algorithmic Guarantees Hamid Reza Feyzmahdavian, Mikael Johansson , 2023. [ abs ][ pdf ][ bib ]

Restarted Nonconvex Accelerated Gradient Descent: No More Polylogarithmic Factor in the in the O(epsilon^(-7/4)) Complexity Huan Li, Zhouchen Lin , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Integrating Random Effects in Deep Neural Networks Giora Simchoni, Saharon Rosset , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Adaptive Data Depth via Multi-Armed Bandits Tavor Baharav, Tze Leung Lai , 2023. [ abs ][ pdf ][ bib ]

Adapting and Evaluating Influence-Estimation Methods for Gradient-Boosted Decision Trees Jonathan Brophy, Zayd Hammoudeh, Daniel Lowd , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Consistent Model-based Clustering using the Quasi-Bernoulli Stick-breaking Process Cheng Zeng, Jeffrey W Miller, Leo L Duan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Selective inference for k-means clustering Yiqun T. Chen, Daniela M. Witten , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Generalization error bounds for multiclass sparse linear classifiers Tomer Levy, Felix Abramovich , 2023. [ abs ][ pdf ][ bib ]

MALib: A Parallel Framework for Population-based Multi-agent Reinforcement Learning Ming Zhou, Ziyu Wan, Hanjing Wang, Muning Wen, Runzhe Wu, Ying Wen, Yaodong Yang, Yong Yu, Jun Wang, Weinan Zhang , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Controlling Wasserstein Distances by Kernel Norms with Application to Compressive Statistical Learning Titouan Vayer, Rémi Gribonval , 2023. [ abs ][ pdf ][ bib ]

Fast Objective & Duality Gap Convergence for Non-Convex Strongly-Concave Min-Max Problems with PL Condition Zhishuai Guo, Yan Yan, Zhuoning Yuan, Tianbao Yang , 2023. [ abs ][ pdf ][ bib ]

Stochastic Optimization under Distributional Drift Joshua Cutler, Dmitriy Drusvyatskiy, Zaid Harchaoui , 2023. [ abs ][ pdf ][ bib ]

Off-Policy Actor-Critic with Emphatic Weightings Eric Graves, Ehsan Imani, Raksha Kumaraswamy, Martha White , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Memory-Based Optimization Methods for Model-Agnostic Meta-Learning and Personalized Federated Learning Bokun Wang, Zhuoning Yuan, Yiming Ying, Tianbao Yang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Escaping The Curse of Dimensionality in Bayesian Model-Based Clustering Noirrit Kiran Chandra, Antonio Canale, David B. Dunson , 2023. [ abs ][ pdf ][ bib ]

Large sample spectral analysis of graph-based multi-manifold clustering Nicolas Garcia Trillos, Pengfei He, Chenghui Li , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On Tilted Losses in Machine Learning: Theory and Applications Tian Li, Ahmad Beirami, Maziar Sanjabi, Virginia Smith , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Optimal Convergence Rates for Distributed Nystroem Approximation Jian Li, Yong Liu, Weiping Wang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Jump Interval-Learning for Individualized Decision Making with Continuous Treatments Hengrui Cai, Chengchun Shi, Rui Song, Wenbin Lu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Policy Gradient Methods Find the Nash Equilibrium in N-player General-sum Linear-quadratic Games Ben Hambly, Renyuan Xu, Huining Yang , 2023. [ abs ][ pdf ][ bib ]

Asymptotics of Network Embeddings Learned via Subsampling Andrew Davison, Morgane Austern , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Implicit Bias of Gradient Descent for Mean Squared Error Regression with Two-Layer Wide Neural Networks Hui Jin, Guido Montufar , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Dimension Reduction in Contextual Online Learning via Nonparametric Variable Selection Wenhao Li, Ningyuan Chen, L. Jeff Hong , 2023. [ abs ][ pdf ][ bib ]

Sparse GCA and Thresholded Gradient Descent Sheng Gao, Zongming Ma , 2023. [ abs ][ pdf ][ bib ]

MARS: A Second-Order Reduction Algorithm for High-Dimensional Sparse Precision Matrices Estimation Qian Li, Binyan Jiang, Defeng Sun , 2023. [ abs ][ pdf ][ bib ]

Exploiting Discovered Regression Discontinuities to Debias Conditioned-on-observable Estimators Benjamin Jakubowski, Sriram Somanchi, Edward McFowland III, Daniel B. Neill , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Generalized Linear Models in Non-interactive Local Differential Privacy with Public Data Di Wang, Lijie Hu, Huanyu Zhang, Marco Gaboardi, Jinhui Xu , 2023. [ abs ][ pdf ][ bib ]

A Rigorous Information-Theoretic Definition of Redundancy and Relevancy in Feature Selection Based on (Partial) Information Decomposition Patricia Wollstadt, Sebastian Schmitt, Michael Wibral , 2023. [ abs ][ pdf ][ bib ]

Combinatorial Optimization and Reasoning with Graph Neural Networks Quentin Cappart, Didier Chételat, Elias B. Khalil, Andrea Lodi, Christopher Morris, Petar Veličković , 2023. [ abs ][ pdf ][ bib ]

A First Look into the Carbon Footprint of Federated Learning Xinchi Qiu, Titouan Parcollet, Javier Fernandez-Marques, Pedro P. B. Gusmao, Yan Gao, Daniel J. Beutel, Taner Topal, Akhil Mathur, Nicholas D. Lane , 2023. [ abs ][ pdf ][ bib ]

An Eigenmodel for Dynamic Multilayer Networks Joshua Daniel Loyal, Yuguo Chen , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Graph Clustering with Graph Neural Networks Anton Tsitsulin, John Palowitch, Bryan Perozzi, Emmanuel Müller , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Euler-Lagrange Analysis of Generative Adversarial Networks Siddarth Asokan, Chandra Sekhar Seelamantula , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Statistical Robustness of Empirical Risks in Machine Learning Shaoyan Guo, Huifu Xu, Liwei Zhang , 2023. [ abs ][ pdf ][ bib ]

HiGrad: Uncertainty Quantification for Online Learning and Stochastic Approximation Weijie J. Su, Yuancheng Zhu , 2023. [ abs ][ pdf ][ bib ]

Benign overfitting in ridge regression Alexander Tsigler, Peter L. Bartlett , 2023. [ abs ][ pdf ][ bib ]

Compute-Efficient Deep Learning: Algorithmic Trends and Opportunities Brian R. Bartoldson, Bhavya Kailkhura, Davis Blalock , 2023. [ abs ][ pdf ][ bib ]

Minimal Width for Universal Property of Deep RNN Chang hoon Song, Geonho Hwang, Jun ho Lee, Myungjoo Kang , 2023. [ abs ][ pdf ][ bib ]

Maximum likelihood estimation in Gaussian process regression is ill-posed Toni Karvonen, Chris J. Oates , 2023. [ abs ][ pdf ][ bib ]

An Annotated Graph Model with Differential Degree Heterogeneity for Directed Networks Stefan Stein, Chenlei Leng , 2023. [ abs ][ pdf ][ bib ]

A Unified Framework for Optimization-Based Graph Coarsening Manoj Kumar, Anurag Sharma, Sandeep Kumar , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Deep linear networks can benignly overfit when shallow ones do Niladri S. Chatterji, Philip M. Long , 2023. [ abs ][ pdf ][ bib ]      [ code ]

SQLFlow: An Extensible Toolkit Integrating DB and AI Jun Zhou, Ke Zhang, Lin Wang, Hua Wu, Yi Wang, ChaoChao Chen , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Learning Good State and Action Representations for Markov Decision Process via Tensor Decomposition Chengzhuo Ni, Yaqi Duan, Munther Dahleh, Mengdi Wang, Anru R. Zhang , 2023. [ abs ][ pdf ][ bib ]

Generalization Bounds for Adversarial Contrastive Learning Xin Zou, Weiwei Liu , 2023. [ abs ][ pdf ][ bib ]

The Implicit Bias of Benign Overfitting Ohad Shamir , 2023. [ abs ][ pdf ][ bib ]

The Hyperspherical Geometry of Community Detection: Modularity as a Distance Martijn Gösgens, Remco van der Hofstad, Nelly Litvak , 2023. [ abs ][ pdf ][ bib ]      [ code ]

FLIP: A Utility Preserving Privacy Mechanism for Time Series Tucker McElroy, Anindya Roy, Gaurab Hore , 2023. [ abs ][ pdf ][ bib ]

A General Theory for Federated Optimization with Asynchronous and Heterogeneous Clients Updates Yann Fraboni, Richard Vidal, Laetitia Kameni, Marco Lorenzi , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Dimensionless machine learning: Imposing exact units equivariance Soledad Villar, Weichi Yao, David W. Hogg, Ben Blum-Smith, Bianca Dumitrascu , 2023. [ abs ][ pdf ][ bib ]

Bayesian Calibration of Imperfect Computer Models using Physics-Informed Priors Michail Spitieris, Ingelin Steinsland , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Risk Bounds for Positive-Unlabeled Learning Under the Selected At Random Assumption Olivier Coudray, Christine Keribin, Pascal Massart, Patrick Pamphile , 2023. [ abs ][ pdf ][ bib ]

Concentration analysis of multivariate elliptic diffusions Lukas Trottner, Cathrine Aeckerle-Willems, Claudia Strauch , 2023. [ abs ][ pdf ][ bib ]

Knowledge Hypergraph Embedding Meets Relational Algebra Bahare Fatemi, Perouz Taslakian, David Vazquez, David Poole , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Intrinsic Gaussian Process on Unknown Manifolds with Probabilistic Metrics Mu Niu, Zhenwen Dai, Pokman Cheung, Yizhu Wang , 2023. [ abs ][ pdf ][ bib ]

Sparse Training with Lipschitz Continuous Loss Functions and a Weighted Group L0-norm Constraint Michael R. Metel , 2023. [ abs ][ pdf ][ bib ]

Learning Optimal Group-structured Individualized Treatment Rules with Many Treatments Haixu Ma, Donglin Zeng, Yufeng Liu , 2023. [ abs ][ pdf ][ bib ]

Inference for Gaussian Processes with Matern Covariogram on Compact Riemannian Manifolds Didong Li, Wenpin Tang, Sudipto Banerjee , 2023. [ abs ][ pdf ][ bib ]

FedLab: A Flexible Federated Learning Framework Dun Zeng, Siqi Liang, Xiangjing Hu, Hui Wang, Zenglin Xu , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Connectivity Matters: Neural Network Pruning Through the Lens of Effective Sparsity Artem Vysogorets, Julia Kempe , 2023. [ abs ][ pdf ][ bib ]

An Analysis of Robustness of Non-Lipschitz Networks Maria-Florina Balcan, Avrim Blum, Dravyansh Sharma, Hongyang Zhang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Fitting Autoregressive Graph Generative Models through Maximum Likelihood Estimation Xu Han, Xiaohui Chen, Francisco J. R. Ruiz, Li-Ping Liu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Global Convergence of Sub-gradient Method for Robust Matrix Recovery: Small Initialization, Noisy Measurements, and Over-parameterization Jianhao Ma, Salar Fattahi , 2023. [ abs ][ pdf ][ bib ]

Statistical Inference for Noisy Incomplete Binary Matrix Yunxiao Chen, Chengcheng Li, Jing Ouyang, Gongjun Xu , 2023. [ abs ][ pdf ][ bib ]

Faith-Shap: The Faithful Shapley Interaction Index Che-Ping Tsai, Chih-Kuan Yeh, Pradeep Ravikumar , 2023. [ abs ][ pdf ][ bib ]

Decentralized Learning: Theoretical Optimality and Practical Improvements Yucheng Lu, Christopher De Sa , 2023. [ abs ][ pdf ][ bib ]

Non-Asymptotic Guarantees for Robust Statistical Learning under Infinite Variance Assumption Lihu Xu, Fang Yao, Qiuran Yao, Huiming Zhang , 2023. [ abs ][ pdf ][ bib ]

Recursive Quantile Estimation: Non-Asymptotic Confidence Bounds Likai Chen, Georg Keilbar, Wei Biao Wu , 2023. [ abs ][ pdf ][ bib ]

Outlier-Robust Subsampling Techniques for Persistent Homology Bernadette J. Stolz , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Neural Operator: Learning Maps Between Function Spaces With Applications to PDEs Nikola Kovachki, Zongyi Li, Burigede Liu, Kamyar Azizzadenesheli, Kaushik Bhattacharya, Andrew Stuart, Anima Anandkumar , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Dimension-Grouped Mixed Membership Models for Multivariate Categorical Data Yuqi Gu, Elena E. Erosheva, Gongjun Xu, David B. Dunson , 2023. [ abs ][ pdf ][ bib ]

Gaussian Processes with Errors in Variables: Theory and Computation Shuang Zhou, Debdeep Pati, Tianying Wang, Yun Yang, Raymond J. Carroll , 2023. [ abs ][ pdf ][ bib ]

Learning Partial Differential Equations in Reproducing Kernel Hilbert Spaces George Stepaniants , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Doubly Robust Stein-Kernelized Monte Carlo Estimator: Simultaneous Bias-Variance Reduction and Supercanonical Convergence Henry Lam, Haofeng Zhang , 2023. [ abs ][ pdf ][ bib ]

Online Optimization over Riemannian Manifolds Xi Wang, Zhipeng Tu, Yiguang Hong, Yingyi Wu, Guodong Shi , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Bayes-Newton Methods for Approximate Bayesian Inference with PSD Guarantees William J. Wilkinson, Simo Särkkä, Arno Solin , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Iterated Block Particle Filter for High-dimensional Parameter Learning: Beating the Curse of Dimensionality Ning Ning, Edward L. Ionides , 2023. [ abs ][ pdf ][ bib ]

Fast Online Changepoint Detection via Functional Pruning CUSUM Statistics Gaetano Romano, Idris A. Eckley, Paul Fearnhead, Guillem Rigaill , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Temporal Abstraction in Reinforcement Learning with the Successor Representation Marlos C. Machado, Andre Barreto, Doina Precup, Michael Bowling , 2023. [ abs ][ pdf ][ bib ]

Approximate Post-Selective Inference for Regression with the Group LASSO Snigdha Panigrahi, Peter W MacDonald, Daniel Kessler , 2023. [ abs ][ pdf ][ bib ]

Towards Learning to Imitate from a Single Video Demonstration Glen Berseth, Florian Golemo, Christopher Pal , 2023. [ abs ][ pdf ][ bib ]

A Likelihood Approach to Nonparametric Estimation of a Singular Distribution Using Deep Generative Models Minwoo Chae, Dongha Kim, Yongdai Kim, Lizhen Lin , 2023. [ abs ][ pdf ][ bib ]

A Randomized Subspace-based Approach for Dimensionality Reduction and Important Variable Selection Di Bo, Hoon Hwangbo, Vinit Sharma, Corey Arndt, Stephanie TerMaath , 2023. [ abs ][ pdf ][ bib ]

Intrinsic Persistent Homology via Density-based Metric Learning Ximena Fernández, Eugenio Borghini, Gabriel Mindlin, Pablo Groisman , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Privacy-Aware Rejection Sampling Jordan Awan, Vinayak Rao , 2023. [ abs ][ pdf ][ bib ]

Inference for a Large Directed Acyclic Graph with Unspecified Interventions Chunlin Li, Xiaotong Shen, Wei Pan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

How Do You Want Your Greedy: Simultaneous or Repeated? Moran Feldman, Christopher Harshaw, Amin Karbasi , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Kernel-Matrix Determinant Estimates from stopped Cholesky Decomposition Simon Bartels, Wouter Boomsma, Jes Frellsen, Damien Garreau , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Optimizing ROC Curves with a Sort-Based Surrogate Loss for Binary Classification and Changepoint Detection Jonathan Hillman, Toby Dylan Hocking , 2023. [ abs ][ pdf ][ bib ]      [ code ]

When Locally Linear Embedding Hits Boundary Hau-Tieng Wu, Nan Wu , 2023. [ abs ][ pdf ][ bib ]

Distributed Nonparametric Regression Imputation for Missing Response Problems with Large-scale Data Ruoyu Wang, Miaomiao Su, Qihua Wang , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Prior Specification for Bayesian Matrix Factorization via Prior Predictive Matching Eliezer de Souza da Silva, Tomasz Kuśmierczyk, Marcelo Hartmann, Arto Klami , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Posterior Contraction for Deep Gaussian Process Priors Gianluca Finocchio, Johannes Schmidt-Hieber , 2023. [ abs ][ pdf ][ bib ]

Wide-minima Density Hypothesis and the Explore-Exploit Learning Rate Schedule Nikhil Iyer, V. Thejas, Nipun Kwatra, Ramachandran Ramjee, Muthian Sivathanu , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Fundamental limits and algorithms for sparse linear regression with sublinear sparsity Lan V. Truong , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On the Complexity of SHAP-Score-Based Explanations: Tractability via Knowledge Compilation and Non-Approximability Results Marcelo Arenas, Pablo Barcelo, Leopoldo Bertossi, Mikael Monet , 2023. [ abs ][ pdf ][ bib ]

Monotonic Alpha-divergence Minimisation for Variational Inference Kamélia Daudel, Randal Douc, François Roueff , 2023. [ abs ][ pdf ][ bib ]

Density estimation on low-dimensional manifolds: an inflation-deflation approach Christian Horvat, Jean-Pascal Pfister , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Provably Sample-Efficient Model-Free Algorithm for MDPs with Peak Constraints Qinbo Bai, Vaneet Aggarwal, Ather Gattami , 2023. [ abs ][ pdf ][ bib ]

Topological Convolutional Layers for Deep Learning Ephy R. Love, Benjamin Filippenko, Vasileios Maroulas, Gunnar Carlsson , 2023. [ abs ][ pdf ][ bib ]

Online Stochastic Gradient Descent with Arbitrary Initialization Solves Non-smooth, Non-convex Phase Retrieval Yan Shuo Tan, Roman Vershynin , 2023. [ abs ][ pdf ][ bib ]

Tree-AMP: Compositional Inference with Tree Approximate Message Passing Antoine Baker, Florent Krzakala, Benjamin Aubin, Lenka Zdeborová , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On the geometry of Stein variational gradient descent Andrew Duncan, Nikolas Nüsken, Lukasz Szpruch , 2023. [ abs ][ pdf ][ bib ]

Kernel-based estimation for partially functional linear model: Minimax rates and randomized sketches Shaogao Lv, Xin He, Junhui Wang , 2023. [ abs ][ pdf ][ bib ]

Contextual Stochastic Block Model: Sharp Thresholds and Contiguity Chen Lu, Subhabrata Sen , 2023. [ abs ][ pdf ][ bib ]

VCG Mechanism Design with Unknown Agent Values under Stochastic Bandit Feedback Kirthevasan Kandasamy, Joseph E Gonzalez, Michael I Jordan, Ion Stoica , 2023. [ abs ][ pdf ][ bib ]

Necessary and Sufficient Conditions for Inverse Reinforcement Learning of Bayesian Stopping Time Problems Kunal Pattanayak, Vikram Krishnamurthy , 2023. [ abs ][ pdf ][ bib ]

Online Change-Point Detection in High-Dimensional Covariance Structure with Application to Dynamic Networks Lingjun Li, Jun Li , 2023. [ abs ][ pdf ][ bib ]

Convergence Rates of a Class of Multivariate Density Estimation Methods Based on Adaptive Partitioning Linxi Liu, Dangna Li, Wing Hung Wong , 2023. [ abs ][ pdf ][ bib ]

Reinforcement Learning for Joint Optimization of Multiple Rewards Mridul Agarwal, Vaneet Aggarwal , 2023. [ abs ][ pdf ][ bib ]

On the Convergence of Stochastic Gradient Descent with Bandwidth-based Step Size Xiaoyu Wang, Ya-xiang Yuan , 2023. [ abs ][ pdf ][ bib ]

A Group-Theoretic Approach to Computational Abstraction: Symmetry-Driven Hierarchical Clustering Haizi Yu, Igor Mineyev, Lav R. Varshney , 2023. [ abs ][ pdf ][ bib ]

The d-Separation Criterion in Categorical Probability Tobias Fritz, Andreas Klingler , 2023. [ abs ][ pdf ][ bib ]

The multimarginal optimal transport formulation of adversarial multiclass classification Nicolás García Trillos, Matt Jacobs, Jakwang Kim , 2023. [ abs ][ pdf ][ bib ]

Robust Load Balancing with Machine Learned Advice Sara Ahmadian, Hossein Esfandiari, Vahab Mirrokni, Binghui Peng , 2023. [ abs ][ pdf ][ bib ]

Benchmarking Graph Neural Networks Vijay Prakash Dwivedi, Chaitanya K. Joshi, Anh Tuan Luu, Thomas Laurent, Yoshua Bengio, Xavier Bresson , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A Simple Approach to Improve Single-Model Deep Uncertainty via Distance-Awareness Jeremiah Zhe Liu, Shreyas Padhy, Jie Ren, Zi Lin, Yeming Wen, Ghassen Jerfel, Zachary Nado, Jasper Snoek, Dustin Tran, Balaji Lakshminarayanan , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Neural Implicit Flow: a mesh-agnostic dimensionality reduction paradigm of spatio-temporal data Shaowu Pan, Steven L. Brunton, J. Nathan Kutz , 2023. [ abs ][ pdf ][ bib ]      [ code ]

On Batch Teaching Without Collusion Shaun Fallat, David Kirkpatrick, Hans U. Simon, Abolghasem Soltani, Sandra Zilles , 2023. [ abs ][ pdf ][ bib ]

Sensing Theorems for Unsupervised Learning in Linear Inverse Problems Julián Tachella, Dongdong Chen, Mike Davies , 2023. [ abs ][ pdf ][ bib ]

First-Order Algorithms for Nonlinear Generalized Nash Equilibrium Problems Michael I. Jordan, Tianyi Lin, Manolis Zampetakis , 2023. [ abs ][ pdf ][ bib ]

Ridges, Neural Networks, and the Radon Transform Michael Unser , 2023. [ abs ][ pdf ][ bib ]

Label Distribution Changing Learning with Sample Space Expanding Chao Xu, Hong Tao, Jing Zhang, Dewen Hu, Chenping Hou , 2023. [ abs ][ pdf ][ bib ]

Can Reinforcement Learning Find Stackelberg-Nash Equilibria in General-Sum Markov Games with Myopically Rational Followers? Han Zhong, Zhuoran Yang, Zhaoran Wang, Michael I. Jordan , 2023. [ abs ][ pdf ][ bib ]

Quantus: An Explainable AI Toolkit for Responsible Evaluation of Neural Network Explanations and Beyond Anna Hedström, Leander Weber, Daniel Krakowczyk, Dilyara Bareeva, Franz Motzkus, Wojciech Samek, Sebastian Lapuschkin, Marina M.-C. Höhne , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Gap Minimization for Knowledge Sharing and Transfer Boyu Wang, Jorge A. Mendez, Changjian Shui, Fan Zhou, Di Wu, Gezheng Xu, Christian Gagné, Eric Eaton , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Sparse PCA: a Geometric Approach Dimitris Bertsimas, Driss Lahlou Kitane , 2023. [ abs ][ pdf ][ bib ]

Labels, Information, and Computation: Efficient Learning Using Sufficient Labels Shiyu Duan, Spencer Chang, Jose C. Principe , 2023. [ abs ][ pdf ][ bib ]

Attacks against Federated Learning Defense Systems and their Mitigation Cody Lewis, Vijay Varadharajan, Nasimul Noman , 2023. [ abs ][ pdf ][ bib ]      [ code ]

HiClass: a Python Library for Local Hierarchical Classification Compatible with Scikit-learn Fábio M. Miranda, Niklas Köhnecke, Bernhard Y. Renard , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Impact of classification difficulty on the weight matrices spectra in Deep Learning and application to early-stopping XuranMeng, JeffYao , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

The SKIM-FA Kernel: High-Dimensional Variable Selection and Nonlinear Interaction Discovery in Linear Time Raj Agrawal, Tamara Broderick , 2023. [ abs ][ pdf ][ bib ]

Generalization Bounds for Noisy Iterative Algorithms Using Properties of Additive Noise Channels Hao Wang, Rui Gao, Flavio P. Calmon , 2023. [ abs ][ pdf ][ bib ]

Discrete Variational Calculus for Accelerated Optimization Cédric M. Campos, Alejandro Mahillo, David Martín de Diego , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Calibrated Multiple-Output Quantile Regression with Representation Learning Shai Feldman, Stephen Bates, Yaniv Romano , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Bayesian Data Selection Eli N. Weinstein, Jeffrey W. Miller , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Lower Bounds and Accelerated Algorithms for Bilevel Optimization Kaiyi ji, Yingbin Liang , 2023. [ abs ][ pdf ][ bib ]

Graph-Aided Online Multi-Kernel Learning Pouya M. Ghari, Yanning Shen , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Interpolating Classifiers Make Few Mistakes Tengyuan Liang, Benjamin Recht , 2023. [ abs ][ pdf ][ bib ]

Regularized Joint Mixture Models Konstantinos Perrakis, Thomas Lartigue, Frank Dondelinger, Sach Mukherjee , 2023. [ abs ][ pdf ][ bib ]      [ code ]

An Inertial Block Majorization Minimization Framework for Nonsmooth Nonconvex Optimization Le Thi Khanh Hien, Duy Nhat Phan, Nicolas Gillis , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Learning Mean-Field Games with Discounted and Average Costs Berkay Anahtarci, Can Deha Kariksiz, Naci Saldi , 2023. [ abs ][ pdf ][ bib ]

Globally-Consistent Rule-Based Summary-Explanations for Machine Learning Models: Application to Credit-Risk Evaluation Cynthia Rudin, Yaron Shaposhnik , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Extending Adversarial Attacks to Produce Adversarial Class Probability Distributions Jon Vadillo, Roberto Santana, Jose A. Lozano , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Python package for causal discovery based on LiNGAM Takashi Ikeuchi, Mayumi Ide, Yan Zeng, Takashi Nicholas Maeda, Shohei Shimizu , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Adaptation to the Range in K-Armed Bandits Hédi Hadiji, Gilles Stoltz , 2023. [ abs ][ pdf ][ bib ]

Learning-augmented count-min sketches via Bayesian nonparametrics Emanuele Dolera, Stefano Favaro, Stefano Peluchetti , 2023. [ abs ][ pdf ][ bib ]

Optimal Strategies for Reject Option Classifiers Vojtech Franc, Daniel Prusa, Vaclav Voracek , 2023. [ abs ][ pdf ][ bib ]

A Line-Search Descent Algorithm for Strict Saddle Functions with Complexity Guarantees Michael J. O'Neill, Stephen J. Wright , 2023. [ abs ][ pdf ][ bib ]

Sampling random graph homomorphisms and applications to network data analysis Hanbaek Lyu, Facundo Memoli, David Sivakoff , 2023. [ abs ][ pdf ][ bib ]      [ code ]

A Relaxed Inertial Forward-Backward-Forward Algorithm for Solving Monotone Inclusions with Application to GANs Radu I. Bot, Michael Sedlmayer, Phan Tu Vuong , 2023. [ abs ][ pdf ][ bib ]

On Distance and Kernel Measures of Conditional Dependence Tianhong Sheng, Bharath K. Sriperumbudur , 2023. [ abs ][ pdf ][ bib ]

AutoKeras: An AutoML Library for Deep Learning Haifeng Jin, François Chollet, Qingquan Song, Xia Hu , 2023. (Machine Learning Open Source Software Paper) [ abs ][ pdf ][ bib ]      [ code ]

Cluster-Specific Predictions with Multi-Task Gaussian Processes Arthur Leroy, Pierre Latouche, Benjamin Guedj, Servane Gey , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Efficient Structure-preserving Support Tensor Train Machine Kirandeep Kour, Sergey Dolgov, Martin Stoll, Peter Benner , 2023. [ abs ][ pdf ][ bib ]      [ code ]

Bayesian Spiked Laplacian Graphs Leo L Duan, George Michailidis, Mingzhou Ding , 2023. [ abs ][ pdf ][ bib ]      [ code ]

The Brier Score under Administrative Censoring: Problems and a Solution Håvard Kvamme, Ørnulf Borgan , 2023. [ abs ][ pdf ][ bib ]

Approximation Bounds for Hierarchical Clustering: Average Linkage, Bisecting K-means, and Local Search Benjamin Moseley, Joshua R. Wang , 2023. [ abs ][ pdf ][ bib ]

machine learning Recently Published Documents

Total documents.

  • Latest Documents
  • Most Cited Documents
  • Contributed Authors
  • Related Sources
  • Related Keywords

An explainable machine learning model for identifying geographical origins of sea cucumber Apostichopus japonicus based on multi-element profile

A comparison of machine learning- and regression-based models for predicting ductility ratio of rc beam-column joints, alexa, is this a historical record.

Digital transformation in government has brought an increase in the scale, variety, and complexity of records and greater levels of disorganised data. Current practices for selecting records for transfer to The National Archives (TNA) were developed to deal with paper records and are struggling to deal with this shift. This article examines the background to the problem and outlines a project that TNA undertook to research the feasibility of using commercially available artificial intelligence tools to aid selection. The project AI for Selection evaluated a range of commercial solutions varying from off-the-shelf products to cloud-hosted machine learning platforms, as well as a benchmarking tool developed in-house. Suitability of tools depended on several factors, including requirements and skills of transferring bodies as well as the tools’ usability and configurability. This article also explores questions around trust and explainability of decisions made when using AI for sensitive tasks such as selection.

Automated Text Classification of Maintenance Data of Higher Education Buildings Using Text Mining and Machine Learning Techniques

Data-driven analysis and machine learning for energy prediction in distributed photovoltaic generation plants: a case study in queensland, australia, modeling nutrient removal by membrane bioreactor at a sewage treatment plant using machine learning models, big five personality prediction based in indonesian tweets using machine learning methods.

<span lang="EN-US">The popularity of social media has drawn the attention of researchers who have conducted cross-disciplinary studies examining the relationship between personality traits and behavior on social media. Most current work focuses on personality prediction analysis of English texts, but Indonesian has received scant attention. Therefore, this research aims to predict user’s personalities based on Indonesian text from social media using machine learning techniques. This paper evaluates several machine learning techniques, including <a name="_Hlk87278444"></a>naive Bayes (NB), K-nearest neighbors (KNN), and support vector machine (SVM), based on semantic features including emotion, sentiment, and publicly available Twitter profile. We predict the personality based on the big five personality model, the most appropriate model for predicting user personality in social media. We examine the relationships between the semantic features and the Big Five personality dimensions. The experimental results indicate that the Big Five personality exhibit distinct emotional, sentimental, and social characteristics and that SVM outperformed NB and KNN for Indonesian. In addition, we observe several terms in Indonesian that specifically refer to each personality type, each of which has distinct emotional, sentimental, and social features.</span>

Compressive strength of concrete with recycled aggregate; a machine learning-based evaluation

Temperature prediction of flat steel box girders of long-span bridges utilizing in situ environmental parameters and machine learning, computer-assisted cohort identification in practice.

The standard approach to expert-in-the-loop machine learning is active learning, where, repeatedly, an expert is asked to annotate one or more records and the machine finds a classifier that respects all annotations made until that point. We propose an alternative approach, IQRef , in which the expert iteratively designs a classifier and the machine helps him or her to determine how well it is performing and, importantly, when to stop, by reporting statistics on a fixed, hold-out sample of annotated records. We justify our approach based on prior work giving a theoretical model of how to re-use hold-out data. We compare the two approaches in the context of identifying a cohort of EHRs and examine their strengths and weaknesses through a case study arising from an optometric research problem. We conclude that both approaches are complementary, and we recommend that they both be employed in conjunction to address the problem of cohort identification in health research.

Export Citation Format

Share document.

  • Published: 17 April 2021

Artificial intelligence and machine learning research: towards digital transformation at a global scale

  • Akila Sarirete 1 ,
  • Zain Balfagih 1 ,
  • Tayeb Brahimi 1 ,
  • Miltiadis D. Lytras 1 , 2 &
  • Anna Visvizi 3 , 4  

Journal of Ambient Intelligence and Humanized Computing volume  13 ,  pages 3319–3321 ( 2022 ) Cite this article

5203 Accesses

9 Citations

Metrics details

Working on a manuscript?

Artificial intelligence (AI) is reshaping how we live, learn, and work. Until recently, AI used to be a fanciful concept, more closely associated with science fiction rather than with anything else. However, driven by unprecedented advances in sophisticated information and communication technology (ICT), AI today is synonymous technological progress already attained and the one yet to come in all spheres of our lives (Chui et al. 2018 ; Lytras et al. 2018 , 2019 ).

Considering that Machine Learning (ML) and AI are apt to reach unforeseen levels of accuracy and efficiency, this special issue sought to promote research on AI and ML seen as functions of data-driven innovation and digital transformation. The combination of expanding ICT-driven capabilities and capacities identifiable across our socio-economic systems along with growing consumer expectations vis-a-vis technology and its value-added for our societies, requires multidisciplinary research and research agenda on AI and ML (Lytras et al. 2021 ; Visvizi et al. 2020 ; Chui et al. 2020 ). Such a research agenda should oscilate around the following five defining issues (Fig. 1 ):

figure 1

Source: The Authors

An AI-Driven Digital Transformation in all aspects of human activity/

Integration of diverse data-warehouses to unified ecosystems of AI and ML value-based services

Deployment of robust AI and ML processing capabilities for enhanced decision making and generation of value our of data.

Design of innovative novel AI and ML applications for predictive and analytical capabilities

Design of sophisticated AI and ML-enabled intelligence components with critical social impact

Promotion of the Digital Transformation in all the aspects of human activity including business, healthcare, government, commerce, social intelligence etc.

Such development will also have a critical impact on government, policies, regulations and initiatives aiming to interpret the value of the AI-driven digital transformation to the sustainable economic development of our planet. Additionally the disruptive character of AI and ML technology and research will required further research on business models and management of innovation capabilities.

This special issue is based on submissions invited from the 17th Annual Learning and Technology Conference 2019 that was held at Effat University and open call jointly. Several very good submissions were received. All of them were subjected a rigorous peer review process specific to the Ambient Intelligence and Humanized Computing Journal.

A variety of innovative topics are included in the agenda of the published papers in this special issue including topics such as:

Stock market Prediction using Machine learning

Detection of Apple Diseases and Pests based on Multi-Model LSTM-based Convolutional Neural Networks

ML for Searching

Machine Learning for Learning Automata

Entity recognition & Relation Extraction

Intelligent Surveillance Systems

Activity Recognition and K-Means Clustering

Distributed Mobility Management

Review Rating Prediction with Deep Learning

Cybersecurity: Botnet detection with Deep learning

Self-Training methods

Neuro-Fuzzy Inference systems

Fuzzy Controllers

Monarch Butterfly Optimized Control with Robustness Analysis

GMM methods for speaker age and gender classification

Regression methods for Permeability Prediction of Petroleum Reservoirs

Surface EMG Signal Classification

Pattern Mining

Human Activity Recognition in Smart Environments

Teaching–Learning based Optimization Algorithm

Big Data Analytics

Diagnosis based on Event-Driven Processing and Machine Learning for Mobile Healthcare

Over a decade ago, Effat University envisioned a timely platform that brings together educators, researchers and tech enthusiasts under one roof and functions as a fount for creativity and innovation. It was a dream that such platform bridges the existing gap and becomes a leading hub for innovators across disciplines to share their knowledge and exchange novel ideas. It was in 2003 that this dream was realized and the first Learning & Technology Conference was held. Up until today, the conference has covered a variety of cutting-edge themes such as Digital Literacy, Cyber Citizenship, Edutainment, Massive Open Online Courses, and many, many others. The conference has also attracted key, prominent figures in the fields of sciences and technology such as Farouq El Baz from NASA, Queen Rania Al-Abdullah of Jordan, and many others who addressed large, eager-to-learn audiences and inspired many with unique stories.

While emerging innovations, such as Artificial Intelligence technologies, are seen today as promising instruments that could pave our way to the future, these were also the focal points around which fruitful discussions have always taken place here at the L&T. The (AI) was selected for this conference due to its great impact. The Saudi government realized this impact of AI and already started actual steps to invest in AI. It is stated in the Kingdome Vision 2030: "In technology, we will increase our investments in, and lead, the digital economy." Dr. Ahmed Al Theneyan, Deputy Minister of Technology, Industry and Digital Capabilities, stated that: "The Government has invested around USD 3 billion in building the infrastructure so that the country is AI-ready and can become a leader in AI use." Vision 2030 programs also promote innovation in technologies. Another great step that our country made is establishing NEOM city (the model smart city).

Effat University realized this ambition and started working to make it a reality by offering academic programs that support the different sectors needed in such projects. For example, the master program in Energy Engineering was launched four years ago to support the energy sector. Also, the bachelor program of Computer Science has tracks in Artificial Intelligence and Cyber Security which was launched in Fall 2020 semester. Additionally, Energy & Technology and Smart Building Research Centers were established to support innovation in the technology and energy sectors. In general, Effat University works effectively in supporting the KSA to achieve its vision in this time of national transformation by graduating skilled citizen in different fields of technology.

The guest editors would like to take this opportunity to thank all the authors for the efforts they put in the preparation of their manuscripts and for their valuable contributions. We wish to express our deepest gratitude to the referees, who provided instrumental and constructive feedback to the authors. We also extend our sincere thanks and appreciation for the organizing team under the leadership of the Chair of L&T 2019 Conference Steering Committee, Dr. Haifa Jamal Al-Lail, University President, for her support and dedication.

Our sincere thanks go to the Editor-in-Chief for his kind help and support.

Chui KT, Lytras MD, Visvizi A (2018) Energy sustainability in smart cities: artificial intelligence, smart monitoring, and optimization of energy consumption. Energies 11(11):2869

Article   Google Scholar  

Chui KT, Fung DCL, Lytras MD, Lam TM (2020) Predicting at-risk university students in a virtual learning environment via a machine learning algorithm. Comput Human Behav 107:105584

Lytras MD, Visvizi A, Daniela L, Sarirete A, De Pablos PO (2018) Social networks research for sustainable smart education. Sustainability 10(9):2974

Lytras MD, Visvizi A, Sarirete A (2019) Clustering smart city services: perceptions, expectations, responses. Sustainability 11(6):1669

Lytras MD, Visvizi A, Chopdar PK, Sarirete A, Alhalabi W (2021) Information management in smart cities: turning end users’ views into multi-item scale development, validation, and policy-making recommendations. Int J Inf Manag 56:102146

Visvizi A, Jussila J, Lytras MD, Ijäs M (2020) Tweeting and mining OECD-related microcontent in the post-truth era: A cloud-based app. Comput Human Behav 107:105958

Download references

Author information

Authors and affiliations.

Effat College of Engineering, Effat Energy and Technology Research Center, Effat University, P.O. Box 34689, Jeddah, Saudi Arabia

Akila Sarirete, Zain Balfagih, Tayeb Brahimi & Miltiadis D. Lytras

King Abdulaziz University, Jeddah, 21589, Saudi Arabia

Miltiadis D. Lytras

Effat College of Business, Effat University, P.O. Box 34689, Jeddah, Saudi Arabia

Anna Visvizi

Institute of International Studies (ISM), SGH Warsaw School of Economics, Aleja Niepodległości 162, 02-554, Warsaw, Poland

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Akila Sarirete .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Cite this article.

Sarirete, A., Balfagih, Z., Brahimi, T. et al. Artificial intelligence and machine learning research: towards digital transformation at a global scale. J Ambient Intell Human Comput 13 , 3319–3321 (2022).

Download citation

Published : 17 April 2021

Issue Date : July 2022


Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

  • Find a journal
  • Publish with us

research paper machine learning

Analytics Insight

Top 10 Machine Learning Research Papers of 2021

' src=

Machine learning research papers showcasing the transformation of the technology

Unbiased gradient estimation in unrolled computation graphs with persistent evolution, solving high-dimensional parabolic pdes using the tensor train format.


Oops I took a gradient: Scalable sampling for discrete distributions

Optimal complexity in decentralized training, understanding self-supervised learning dynamics without contrastive pairs, how transferable are featured in deep neural networks, do we need hundreds of classifiers to solve real-world classification problems, knowledge vault: a web-scale approach to probabilistic knowledge fusion, scalable nearest neighbor algorithms for high dimensional data, trends in extreme learning machines.

research paper machine learning

Disclaimer: Any financial and crypto market information given on Analytics Insight are sponsored articles, written for informational purpose only and is not an investment advice. The readers are further advised that Crypto products and NFTs are unregulated and can be highly risky. There may be no regulatory recourse for any loss from such transactions. Conduct your own research by contacting financial experts before making any investment decisions. The decision to read hereinafter is purely a matter of choice and shall be construed as an express undertaking/guarantee in favour of Analytics Insight of being absolved from any/ all potential legal action, or enforceable claims. We do not represent nor own any cryptocurrency, any complaints, abuse or concerns with regards to the information provided shall be immediately informed here .

You May Also Like

Cryptocurrency market

Should I Buy Shiba Inu In 2022? Pros And Cons Of Shiba Inu Investment

research paper machine learning

Measuring Organization’s Cybersecurity

Tech Stocks

Consider Top 5 Tech Stocks on November 1, 2021 to Yield Profit in Future

Top ChatGpt Predictions

Top ChatGpt Predictions of 2023

research paper machine learning

Analytics Insight® is an influential platform dedicated to insights, trends, and opinion from the world of data-driven technologies. It monitors developments, recognition, and achievements made by Artificial Intelligence, Big Data and Analytics companies across the globe.


  • Select Language:
  • Privacy Policy
  • Content Licensing
  • Terms & Conditions
  • Submit an Interview

Special Editions

  • 40 Under 40 Innovators
  • Women In Technology
  • Market Reports
  • AI Glossary
  • Infographics

Latest Issue

Crypto Weekly November 2023

Disclaimer: Any financial and crypto market information given on Analytics Insight is written for informational purpose only and is not an investment advice. Conduct your own research by contacting financial experts before making any investment decisions, more information here .

Second Menu

research paper machine learning

Machine Learning

Ieee account.

  • Change Username/Password
  • Update Address

Purchase Details

  • Payment Options
  • Order History
  • View Purchased Documents

Profile Information

  • Communications Preferences
  • Profession and Education
  • Technical Interests
  • US & Canada: +1 800 678 4333
  • Worldwide: +1 732 981 0060
  • Contact & Support
  • About IEEE Xplore
  • Accessibility
  • Terms of Use
  • Nondiscrimination Policy
  • Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2023 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

Help | Advanced Search

Computer Science > Machine Learning

Title: jaxmarl: multi-agent rl environments in jax.

Abstract: Benchmarks play an important role in the development of machine learning algorithms. For example, research in reinforcement learning (RL) has been heavily influenced by available environments and benchmarks. However, RL environments are traditionally run on the CPU, limiting their scalability with typical academic compute. Recent advancements in JAX have enabled the wider use of hardware acceleration to overcome these computational hurdles, enabling massively parallel RL training pipelines and environments. This is particularly useful for multi-agent reinforcement learning (MARL) research. First of all, multiple agents must be considered at each environment step, adding computational burden, and secondly, the sample complexity is increased due to non-stationarity, decentralised partial observability, or other MARL challenges. In this paper, we present JaxMARL, the first open-source code base that combines ease-of-use with GPU enabled efficiency, and supports a large number of commonly used MARL environments as well as popular baseline algorithms. When considering wall clock time, our experiments show that per-run our JAX-based training pipeline is up to 12500x faster than existing approaches. This enables efficient and thorough evaluations, with the potential to alleviate the evaluation crisis of the field. We also introduce and benchmark SMAX, a vectorised, simplified version of the popular StarCraft Multi-Agent Challenge, which removes the need to run the StarCraft II game engine. This not only enables GPU acceleration, but also provides a more flexible MARL environment, unlocking the potential for self-play, meta-learning, and other future applications in MARL. We provide code at this https URL .

Submission history

Access paper:.

  • Download PDF
  • Other Formats

research paper machine learning

References & Citations

  • Google Scholar
  • Semantic Scholar

BibTeX formatted citation

BibSonomy logo

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

  • Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

Publication database

Publishing our work allows us to share ideas and work collaboratively to advance the field of computer science.

Filters Reset filters

Algorithms and theory.

Google’s mission presents many exciting algorithmic and optimization challenges across different product areas including Search, Ads, Social, and Google Infrastructure. These include optimizing internal systems such as scheduling the machines that power the numerous computations done each day, as well as optimizations that affect core products and users, from online allocation of ads to page-views to automatic management of ad campaigns, and from clustering large-scale graphs to finding best paths in transportation networks. Other than employing new algorithmic ideas to impact millions of users, Google researchers contribute to the state-of-the-art research in these areas by publishing in top conferences and journals.

Data Management

Google is deeply engaged in Data Management research across a variety of topics with deep connections to Google products. We are building intelligent systems to discover, annotate, and explore structured data from the Web, and to surface them creatively through Google products, such as Search (e.g., structured snippets , Docs, and many others). The overarching goal is to create a plethora of structured data on the Web that maximally help Google users consume, interact and explore information. Through those projects, we study various cutting-edge data management research issues including information extraction and integration, large scale data analysis, effective data exploration, etc., using a variety of techniques, such as information retrieval, data mining and machine learning.

A major research effort involves the management of structured data within the enterprise. The goal is to discover, index, monitor, and organize this type of data in order to make it easier to access high-quality datasets. This type of data carries different, and often richer, semantics than structured data on the Web, which in turn raises new opportunities and technical challenges in their management.

Furthermore, Data Management research across Google allows us to build technologies that power Google's largest businesses through scalable, reliable, fast, and general-purpose infrastructure for large-scale data processing as a service. Some examples of such technologies include F1 , the database serving our ads infrastructure; Mesa , a petabyte-scale analytic data warehousing system; and Dremel , for petabyte-scale data processing with interactive response times. Dremel is available for external customers to use as part of Google Cloud’s BigQuery .

Data Mining and Modeling

The proliferation of machine learning means that learned classifiers lie at the core of many products across Google. However, questions in practice are rarely so clean as to just to use an out-of-the-box algorithm. A big challenge is in developing metrics, designing experimental methodologies, and modeling the space to create parsimonious representations that capture the fundamentals of the problem. These problems cut across Google’s products and services, from designing experiments for testing new auction algorithms to developing automated metrics to measure the quality of a road map.

Data mining lies at the heart of many of these questions, and the research done at Google is at the forefront of the field. Whether it is finding more efficient algorithms for working with massive data sets, developing privacy-preserving methods for classification, or designing new machine learning approaches, our group continues to push the boundary of what is possible.

Distributed Systems and Parallel Computing

No matter how powerful individual computers become, there are still reasons to harness the power of multiple computational units, often spread across large geographic areas. Sometimes this is motivated by the need to collect data from widely dispersed locations (e.g., web pages from servers, or sensors for weather or traffic). Other times it is motivated by the need to perform enormous computations that simply cannot be done by a single CPU.

From our company’s beginning, Google has had to deal with both issues in our pursuit of organizing the world’s information and making it universally accessible and useful. We continue to face many exciting distributed systems and parallel computing challenges in areas such as concurrency control, fault tolerance, algorithmic efficiency, and communication. Some of our research involves answering fundamental theoretical questions, while other researchers and engineers are engaged in the construction of systems to operate at the largest possible scale, thanks to our hybrid research model .

Economics and Electronic Commerce

Google is a global leader in electronic commerce. Not surprisingly, it devotes considerable attention to research in this area. Topics include 1) auction design, 2) advertising effectiveness, 3) statistical methods, 4) forecasting and prediction, 5) survey research, 6) policy analysis and a host of other topics. This research involves interdisciplinary collaboration among computer scientists, economists, statisticians, and analytic marketing researchers both at Google and academic institutions around the world.

A major challenge is in solving these problems at very large scales. For example, the advertising market has billions of transactions daily, spread across millions of advertisers. It presents a unique opportunity to test and refine economic principles as applied to a very large number of interacting, self-interested parties with a myriad of objectives.

It is remarkable how some of the fundamental problems Google grapples with are also some of the hardest research problems in the academic community. At Google, this research translates direction into practice, influencing how production systems are designed and used.

Education Innovation

Our Education Innovation research area includes publications on: online learning at scale, educational technology (which is any technology that supports teaching and learning), curriculum and programming tools for computer science education, diversity and broadening participation in computer science the hiring and onboarding process at Google.

General Science

We aim to transform scientific research itself. Many scientific endeavors can benefit from large scale experimentation, data gathering, and machine learning (including deep learning). We aim to accelerate scientific research by applying Google’s computational power and techniques in areas such as drug discovery, biological pathway modeling, microscopy, medical diagnostics, material science, and agriculture. We collaborate closely with world-class research partners to help solve important problems with large scientific or humanitarian benefit.

Hardware and Architecture

The machinery that powers many of our interactions today — Web search, social networking, email, online video, shopping, game playing — is made of the smallest and the most massive computers. The smallest part is your smartphone, a machine that is over ten times faster than the iconic Cray-1 supercomputer. The capabilities of these remarkable mobile devices are amplified by orders of magnitude through their connection to Web services running on building-sized computing systems that we call Warehouse-scale computers (WSCs).

Google’s engineers and researchers have been pioneering both WSC and mobile hardware technology with the goal of providing Google programmers and our Cloud developers with a unique computing infrastructure in terms of scale, cost-efficiency, energy-efficiency, resiliency and speed. The tight collaboration among software, hardware, mechanical, electrical, environmental, thermal and civil engineers result in some of the most impressive and efficient computers in the world.

Human-Computer Interaction and Visualization

HCI researchers at Google have enormous potential to impact the experience of Google users as well as conduct innovative research. Grounded in user behavior understanding and real use, Google’s HCI researchers invent, design, build and trial large-scale interactive systems in the real world. We declare success only when we positively impact our users and user communities, often through new and improved Google products. HCI research has fundamentally contributed to the design of Search, Gmail, Docs, Maps, Chrome, Android, YouTube, serving over a billion daily users. We are engaged in a variety of HCI disciplines such as predictive and intelligent user interface technologies and software, mobile and ubiquitous computing, social and collaborative computing, interactive visualization and visual analytics. Many projects heavily incorporate machine learning with HCI, and current projects include predictive user interfaces; recommenders for content, apps, and activities; smart input and prediction of text on mobile devices; user engagement analytics; user interface development tools; and interactive visualization of complex data.

Information Retrieval and the Web

The science surrounding search engines is commonly referred to as information retrieval, in which algorithmic principles are developed to match user interests to the best information about those interests.

Google started as a result of our founders' attempt to find the best matching between the user queries and Web documents, and do it really fast. During the process, they uncovered a few basic principles: 1) best pages tend to be those linked to the most; 2) best description of a page is often derived from the anchor text associated with the links to a page. Theories were developed to exploit these principles to optimize the task of retrieving the best documents for a user query.

Search and Information Retrieval on the Web has advanced significantly from those early days: 1) the notion of "information" has greatly expanded from documents to much richer representations such as images, videos, etc., 2) users are increasingly searching on their Mobile devices with very different interaction characteristics from search on the Desktops; 3) users are increasingly looking for direct information, such as answers to a question, or seeking to complete tasks, such as appointment booking. Through our research, we are continuing to enhance and refine the world's foremost search engine by aiming to scientifically understand the implications of those changes and address new challenges that they bring.

Machine Intelligence

Google is at the forefront of innovation in Machine Intelligence, with active research exploring virtually all aspects of machine learning, including deep learning and more classical algorithms. Exploring theory as well as application, much of our work on language, speech, translation, visual processing, ranking and prediction relies on Machine Intelligence. In all of those tasks and many others, we gather large volumes of direct or indirect evidence of relationships of interest, applying learning algorithms to understand and generalize.

Machine Intelligence at Google raises deep scientific and engineering challenges, allowing us to contribute to the broader academic research community through technical talks and publications in major conferences and journals. Contrary to much of current theory and practice, the statistics of the data we observe shifts rapidly, the features of interest change as well, and the volume of data often requires enormous computation capacity. When learning systems are placed at the core of interactive services in a fast changing and sometimes adversarial environment, combinations of techniques including deep learning and statistical models need to be combined with ideas from control and game theory.

Machine Perception

Research in machine perception tackles the hard problems of understanding images, sounds, music and video. In recent years, our computers have become much better at such tasks, enabling a variety of new applications such as: content-based search in Google Photos and Image Search, natural handwriting interfaces for Android, optical character recognition for Google Drive documents, and recommendation systems that understand music and YouTube videos. Our approach is driven by algorithms that benefit from processing very large, partially-labeled datasets using parallel computing clusters. A good example is our recent work on object recognition using a novel deep convolutional neural network architecture known as Inception that achieves state-of-the-art results on academic benchmarks and allows users to easily search through their large collection of Google Photos. The ability to mine meaningful information from multimedia is broadly applied throughout Google.

Machine Translation

Machine Translation is an excellent example of how cutting-edge research and world-class infrastructure come together at Google. We focus our research efforts on developing statistical translation techniques that improve with more data and generalize well to new languages. Our large scale computing infrastructure allows us to rapidly experiment with new models trained on web-scale data to significantly improve translation quality. This research backs the translations served at, allowing our users to translate text, web pages and even speech. Deployed within a wide range of Google services like GMail , Books , Android and web search , Google Translate is a high-impact, research-driven product that bridges language barriers and makes it possible to explore the multilingual web in 90 languages. Exciting research challenges abound as we pursue human quality translation and develop machine translation systems for new languages.

Mobile Systems

Mobile devices are the prevalent computing device in many parts of the world, and over the coming years it is expected that mobile Internet usage will outpace desktop usage worldwide. Google is committed to realizing the potential of the mobile web to transform how people interact with computing technology. Google engineers and researchers work on a wide range of problems in mobile computing and networking, including new operating systems and programming platforms (such as Android and ChromeOS); new interaction paradigms between people and devices; advanced wireless communications; and optimizing the web for mobile settings. In addition, many of Google’s core product teams, such as Search, Gmail, and Maps, have groups focused on optimizing the mobile experience, making it faster and more seamless. We take a cross-layer approach to research in mobile systems and networking, cutting across applications, networks, operating systems, and hardware. The tremendous scale of Google’s products and the Android and Chrome platforms make this a very exciting place to work on these problems.

Some representative projects include mobile web performance optimization, new features in Android to greatly reduce network data usage and energy consumption; new platforms for developing high performance web applications on mobile devices; wireless communication protocols that will yield vastly greater performance over today’s standards; and multi-device interaction based on Android, which is now available on a wide variety of consumer electronics.

Natural Language Processing

Natural Language Processing (NLP) research at Google focuses on algorithms that apply at scale, across languages, and across domains. Our systems are used in numerous ways across Google, impacting user experience in search, mobile, apps, ads, translate and more.

Our work spans the range of traditional NLP tasks, with general-purpose syntax and semantic algorithms underpinning more specialized systems. We are particularly interested in algorithms that scale well and can be run efficiently in a highly distributed environment.

Our syntactic systems predict part-of-speech tags for each word in a given sentence, as well as morphological features such as gender and number. They also label relationships between words, such as subject, object, modification, and others. We focus on efficient algorithms that leverage large amounts of unlabeled data, and recently have incorporated neural net technology.

On the semantic side, we identify entities in free text, label them with types (such as person, location, or organization), cluster mentions of those entities within and across documents (coreference resolution), and resolve the entities to the Knowledge Graph.

Recent work has focused on incorporating multiple sources of knowledge and information to aid with analysis of text, as well as applying frame semantics at the noun phrase, sentence, and document level.

Networking is central to modern computing, from connecting cell phones to massive Cloud-based data stores to the interconnect for data centers that deliver seamless storage and fine-grained distributed computing at the scale of entire buildings. With an understanding that our distributed computing infrastructure is a key differentiator for the company, Google has long focused on building network infrastructure to support our scale, availability, and performance needs.

Our research combines building and deploying novel networking systems at massive scale, with recent work focusing on fundamental questions around data center architecture, wide area network interconnects, Software Defined Networking control and management infrastructure, as well as congestion control and bandwidth allocation. By publishing our findings at premier research venues, we continue to engage both academic and industrial partners to further the state of the art in networked systems.

Quantum Computing

Quantum Computing merges two great scientific revolutions of the 20th century: computer science and quantum physics. Quantum physics is the theoretical basis of the transistor, the laser, and other technologies which enabled the computing revolution. But on the algorithmic level, today's computing machinery still operates on "classical" Boolean logic. Quantum Computing is the design of hardware and software that replaces Boolean logic by quantum law at the algorithmic level. For certain computations such as optimization, sampling, search or quantum simulation this promises dramatic speedups. We are particularly interested in applying quantum computing to artificial intelligence and machine learning. This is because many tasks in these areas rely on solving hard optimization problems or performing efficient sampling.

Having a machine learning agent interact with its environment requires true unsupervised learning, skill acquisition, active learning, exploration and reinforcement, all ingredients of human learning that are still not well understood or exploited through the supervised approaches that dominate deep learning today.

Our goal is to improve robotics via machine learning, and improve machine learning via robotics. We foster close collaborations between machine learning researchers and roboticists to enable learning at scale on real and simulated robotic systems.

Security, Privacy and Abuse Prevention

The Internet and the World Wide Web have brought many changes that provide huge benefits, in particular by giving people easy access to information that was previously unavailable, or simply hard to find. Unfortunately, these changes have raised many new challenges in the security of computer systems and the protection of information against unauthorized access and abusive usage. At Google, our primary focus is the user, and his/her safety. We have people working on nearly every aspect of security, privacy, and anti-abuse including access control and information security, networking, operating systems, language design, cryptography, fraud detection and prevention, spam and abuse detection, denial of service, anonymity, privacy-preserving systems, disclosure controls, as well as user interfaces and other human-centered aspects of security and privacy. Our security and privacy efforts cover a broad range of systems including mobile, cloud, distributed, sensors and embedded systems, and large-scale machine learning.

Software Engineering

At Google, we pride ourselves on our ability to develop and launch new products and features at a very fast pace. This is made possible in part by our world-class engineers, but our approach to software development enables us to balance speed and quality, and is integral to our success. Our obsession for speed and scale is evident in our developer infrastructure and tools. Developers across the world continually write, build, test and release code in multiple programming languages like C++, Java, Python, Javascript and others, and the Engineering Tools team, for example, is challenged to keep this development ecosystem running smoothly. Our engineers leverage these tools and infrastructure to produce clean code and keep software development running at an ever-increasing scale. In our publications, we share associated technical challenges and lessons learned along the way.

Software Systems

Delivering Google's products to our users requires computer systems that have a scale previously unknown to the industry. Building on our hardware foundation, we develop technology across the entire systems stack, from operating system device drivers all the way up to multi-site software systems that run on hundreds of thousands of computers. We design, build and operate warehouse-scale computer systems that are deployed across the globe. We build storage systems that scale to exabytes, approach the performance of RAM, and never lose a byte. We design algorithms that transform our understanding of what is possible. Thanks to the distributed systems we provide our developers, they are some of the most productive in the industry. And we write and publish research papers to share what we have learned, and because peer feedback and interaction helps us build better systems that benefit everybody.

Speech Processing

Our goal in Speech Technology Research is to make speaking to devices--those around you, those that you wear, and those that you carry with you--ubiquitous and seamless.

Our research focuses on what makes Google unique: computing scale and data. Using large scale computing resources pushes us to rethink the architecture and algorithms of speech recognition, and experiment with the kind of methods that have in the past been considered prohibitively expensive. We also look at parallelism and cluster computing in a new light to change the way experiments are run, algorithms are developed and research is conducted. The field of speech recognition is data-hungry, and using more and more data to tackle a problem tends to help performance but poses new challenges: how do you deal with data overload? How do you leverage unsupervised and semi-supervised techniques at scale? Which class of algorithms merely compensate for lack of data and which scale well with the task at hand? Increasingly, we find that the answers to these questions are surprising, and steer the whole field into directions that would never have been considered, were it not for the availability of significantly higher orders of magnitude of data.

We are also in a unique position to deliver very user-centric research. Researchers are able to conduct live experiments to test and benchmark new algorithms directly in a realistic controlled environment. Whether these are algorithmic performance improvements or user experience and human-computer interaction studies, we focus on solving real problems and with real impact for users.

We have a huge commitment to the diversity of our users, and have made it a priority to deliver the best performance to every language on the planet. We currently have systems operating in more than 55 languages, and we continue to expand our reach to more users. The challenges of internationalizing at scale is immense and rewarding. Many speakers of the languages we reach have never had the experience of speaking to a computer before, and breaking this new ground brings up new research on how to better serve this wide variety of users. Combined with the unprecedented translation capabilities of Google Translate, we are now at the forefront of research in speech-to-speech translation and one step closer to a universal translator.

Indexing and transcribing the web’s audio content is another challenge we have set for ourselves, and is nothing short of gargantuan, both in scope and difficulty. The videos uploaded every day on YouTube range from lectures, to newscasts, music videos and, of course, cat videos. Making sense of them takes the challenges of noise robustness, music recognition, speaker segmentation, language detection to new levels of difficulty. The potential payoff is immense: imagine making every lecture on the web accessible to every language. This is the kind of impact for which we are striving.

Health & Bioscience

Research in health and biomedical sciences has a unique potential to improve peoples’ lives, and includes work ranging from basic science that aims to understand biology, to diagnosing individuals’ diseases, to epidemiological studies of whole populations.

We recognize that our strengths in machine learning, large-scale computing, and human-computer interaction can help accelerate the progress of research in this space. By collaborating with world-class institutions and researchers and engaging in both early-stage research and late-stage work, we hope to help people live healthier, longer, and more productive lives.

Responsible AI


/// /// ,

Learn more about how we do research

We maintain a portfolio of research projects, providing individuals and teams the freedom to emphasize specific types of work

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • View all journals
  • Explore content
  • About the journal
  • Publish with us
  • Sign up for alerts
  • Published: 13 November 2023

Automated synthesis of oxygen-producing catalysts from Martian meteorites by a robotic AI chemist

  • Qing Zhu   ORCID: 1   na1 ,
  • Yan Huang   ORCID: 1   na1 ,
  • Donglai Zhou 1   na1 ,
  • Luyuan Zhao 1   na1 ,
  • Lulu Guo 1 ,
  • Ruyu Yang 1 ,
  • Zixu Sun   ORCID: 1 ,
  • Man Luo 1 ,
  • Fei Zhang 2 ,
  • Hengyu Xiao 1 ,
  • Xinsheng Tang 2 ,
  • Xuchun Zhang 2 ,
  • Tao Song 2 ,
  • Xiang Li 2 ,
  • Baochen Chong 2 ,
  • Junyi Zhou 2 ,
  • Yihan Zhang 2 ,
  • Baicheng Zhang   ORCID: 1 ,
  • Jiaqi Cao 1 ,
  • Guozhen Zhang   ORCID: 1 ,
  • Song Wang 1 ,
  • Guilin Ye 3 ,
  • Wanjun Zhang 3 ,
  • Haitao Zhao   ORCID: 4 ,
  • Shuang Cong   ORCID: 2 ,
  • Huirong Li 1 ,
  • Li-Li Ling 5 ,
  • Zhe Zhang 5 , 6 ,
  • Weiwei Shang   ORCID: 2 ,
  • Jun Jiang   ORCID: 1 , 7 &
  • Yi Luo   ORCID: 1 , 7  

Nature Synthesis ( 2023 ) Cite this article

5589 Accesses

1 Citations

790 Altmetric

Metrics details

  • Computational methods
  • Electrocatalysis
  • Materials chemistry

Living on Mars requires the ability to synthesize chemicals that are essential for survival, such as oxygen, from local Martian resources. However, this is a challenging task. Here we demonstrate a robotic artificial-intelligence chemist for automated synthesis and intelligent optimization of catalysts for the oxygen evolution reaction from Martian meteorites. The entire process, including Martian ore pretreatment, catalyst synthesis, characterization, testing and, most importantly, the search for the optimal catalyst formula, is performed without human intervention. Using a machine-learning model derived from both first-principles data and experimental measurements, this method automatically and rapidly identifies the optimal catalyst formula from more than three million possible compositions. The synthesized catalyst operates at a current density of 10 mA cm −2 for over 550,000 s of operation with an overpotential of 445.1 mV, demonstrating the feasibility of the artificial-intelligence chemist in the automated synthesis of chemicals and materials for Mars exploration.

research paper machine learning

You have full access to this article via your institution.

Mars has for decades attracted intensive scientific exploration and research in countries worldwide. Finding signs of past life and building potentially habitable regions on Mars have long been a dream of humanity. In situ resource utilization on Mars will be applied to substantially reduce the cost and complexity of human missions, enabling sustainable exploration by utilizing local resources to produce necessary supplies. Oxygen supply must be the top priority for any human activity on Mars, because rocket propellants and life support systems consume substantial amounts of oxygen, which cannot be replenished from the Martian atmosphere 1 , 2 . Fortunately, recent evidence of water activity 3 , 4 has raised the prospect of large-scale oxygen production on the planet through solar-power-driven electrochemical water oxidation processes using an oxygen evolution reaction (OER) catalyst. Using extraterrestrial catalysts developed from local materials to drive oxygen production allows for the on-site production of fuel and oxygen on Mars, which represents a low-hanging fruit in the exploration of this planet. However, two major technical challenges must be overcome to synthesize usable OER catalysts by using local Martian raw materials 5 , 6 . First, the synthetic system must be unmanned and self-directing, as the vast astronomical distance hinders real-time remote guidance from humans. Second, it should be equipped with the scientific intelligence needed to efficiently identify the best formula of catalyst ingredients through artificial-intelligence (AI) algorithms, given knowledge of elemental abundances in the Martian local ores. Designing a catalyst from a given list of elements requires the exploration of a vast chemical space, which poses a daunting task using the conventional ‘trial-and-error’ paradigm. Given five different local Martian ores as feedstocks, there are 3,764,376 possible formulas, estimated by the combination of integer percentages in 1% intervals; finding the optimal formula would require 2,000 years of human labour to finish such a screening, where each complete experiment takes 5 hours, at least.

Robotic synthetic systems with AI appear to be the only viable technology for addressing these two challenges, as suggested by recent advances in automated chemical synthesis systems. The mobile chemist by Cooper and colleagues shows excellent ability to perform high-throughput performance testing for human-made photo-catalysts, providing local optimization with measured data to achieve better formulations 7 . The ChemPU system by Cronin and colleagues demonstrates its extraordinary power in automatic synthesis of organic molecules starting from machine-reading synthetic chemistry literature 8 . These robotic systems need an intelligent subsystem to acquire chemical knowledge and form predictive physical models to direct local optimization in chemical synthesis. Inspired by these researchers' pioneering work on robotic chemical synthesis systems, we have developed an all-in-one robotic artificial-intelligence chemist (AI chemist) to enable automated, self-directed synthesis. Not only can it conduct the entire process of chemical synthesis, structural characterization and performance testing using a mobile robot and 14 task-specific chemistry workstations but it can design the best formula for a chemical synthesis task through a powerful computational module that combines machine learning (ML) algorithms and theoretical models to analyse both robot-acquired experimental data and massive first-principles simulation data 9 . Our AI chemist has accelerated the discovery of the optimal synthetic formulas for high-entropy electrocatalysts by five orders of magnitude compared to conventional trial-and-error experiment paradigm. Without prior knowledge about the exact composition of available Martian ores for making OER catalysts, the proposed automated approach must not only be capable of screening numerous candidates for the best formula, but also be intelligent to dissect usable yet unidentified raw materials and determine the predictive model on-the-fly. We developed a specific protocol for our AI-chemist system to tackle this challenge, advancing the in situ resource utilization strategy for Mars and interstellar exploration in the future.

In this proof-of-concept work, we demonstrate the superiority of the data-driven protocol using an AI chemist over the conventional trial-and-error protocol by the design of a six-metallic element OER catalyst from a pool of 3,764,376 compositions. Within six weeks, the AI chemist built a predictive model by learning from nearly 30,000 theoretical datasets and 243 experimental datasets using ML and Bayesian optimization algorithms, which delivers a promising OER catalyst formula coupled with the most suitable synthetic conditions. The resulting polymetallic material (comprising Mn, Fe, Ni, Mg, Al and Ca) catalysed the OER with an overpotential of 445.1 mV at a current density of 10 mA cm −2 , maintained for 550,000 s. Further, the stress test at −37 °C, which mimics the temperature condition on Mars, confirmed that it can steadily produce oxygen without apparent deterioration, suggesting that it can work in the harsh conditions on Mars. A ground-based verification system is currently being developed to provide more realistic space conditions for the AI chemist, which will be essential for the construction of the International Lunar Research Station and Mars Research Station; both were designed for long-term robotic operation and short-term human participation. The AI chemist thus represents a promising technique for on-site synthesis of OER electrocatalysts on Mars and constitutes a versatile and efficient platform for the supply of complex functional materials for planetary and space exploration.

Protocol for the AI chemist making OER electrocatalysts on Mars

To facilitate the work of the AI chemist on Mars, we proposed a double-layer workflow for the on-site synthesis of OER electrocatalysts (Fig. 1 ). The outer layer, which comprises a 12-step automated experiment and data management, is done by the robot and various ‘smart’ chemical workstations; the inner layer, which includes nine consecutive digital operations, is executed by the intelligent computational ‘brain’ (Supplementary Video 1 and Supplementary Figs. 1 – 3 ).

figure 1

The dual-cycle automated process integrates material preparation, catalyst production, performance characterization and formula optimization in the following steps, as labelled. Step 1: Analyse the precise composition of Martian ores by LIBS. Step 2: Generate polymetallic catalyst structures by classical MD simulations. Step 3: Calculate the OER activities of the structures using DFT. Step 4: Build an NN model using simulation data. Step 5: Re-train the NN model using robotic experimental data. Step 6: Fine-tune the parameters in the NN model to predict experimental overpotential with confidence level over 0.95. Step 7: Screen for the optimum formula using Bayesian optimization algorithms. Step 8: Predict the optimal synthetic formula with the lowest OER overpotential using available Martian ores. Step 9: Validate the OER performance of the catalyst prepared with predicted formula (arrow pointing back to ‘Martian Ore’ box for feedstocks configuration).

In the experimental cycle, samples of local ore (Supplementary Figs. 4 – 9 ) obtained by an exploratory robot are sent to the laser-induced breakdown spectroscopy (LIBS) facility for elemental analysis (Supplementary Fig. 10 ). The robot carries out a set of physical and chemical pretreatment of ores needed for catalyst synthesis, including weighing (with a precision of 0.1 mg) in the solid-dispensing workstation, preparation of feedstock solutions in the liquid-dispensing and mixing workstations (Supplementary Table 1 ), separation from liquid in the centrifugation workstation and solidification in the dryer workstation. Then, the catalyst ink prepared by adding Nafion adhesive into the resulting metal hydroxides is used for making the working electrode for electrochemical OER testing at the electrochemical workstation. Experimental data are sent to a cloud server for ML processing by the computational ‘brain’.

In the computational cycle, the ‘brain’ employs molecular dynamics (MD) simulations for tens of thousands of high-entropy hydroxides with different elemental ratios and applies density functional theory (DFT) calculations to estimate OER activities. Simulation data are used to train a theory-based neural network (NN) model for OER with varying elemental composition, which is soon re-trained and optimized with robot-driven experimental data. By embedding the optimized NN model in a Bayesian algorithm, the ‘brain’ predicts the best combination of available Martian ores for synthesizing an optimal OER catalyst, which is then verified experimentally by the AI chemist.

Building pretrained ML models using the computational ‘brain’

The OER is a thermodynamically uphill reaction involving four consecutive oxidation steps and O–O bond formation, which requires an applied voltage of no less than 1.23 V to operate. The OER overpotential, which is defined as the extra voltage above 1.23 V required for catalysis to occur, characterizes the voltage efficiency of the electrochemical device. Therefore, we chose the measured overpotential as the primary target of our ML model in searching for the optimal OER catalyst 10 , 11 , 12 . We first created 29,902 unique compositions and simulated atomic structures of resulting high-entropy hydroxides (Fig. 2a ) from classical MD simulations (Supplementary Figs. 11 and 12 ). The obtained structural features, such as averaged metal–metal and metal–oxygen distances (Supplementary Figs. 13 and 14 and Supplementary Table 2 ), are passed to previously established bimetallic hydroxide models 13 , 14 (Fig. 2b ) to determine the OER activity of each multimetallic hydroxide by DFT calculation. Three DFT-predicted OER activity descriptors—including the Gibbs free energy change of hydroxyl adsorption Δ G OH* (ref. 15 ) and differences between the Gibbs free energy change for oxygen adsorption and hydroxyl adsorption Δ G O*−OH* (ref. 16 ), the amount of charge transferred for hydroxyl adsorption on the activate site Δ q (ref. 17 )—and the paired composition information are used for NN training. As Fig. 2c shows, the NN model can accurately reproduce these DFT results. With the NN model, we can now rapidly predict the OER activity of high-entropy hydroxides obtained from any given composition of selected Martian ores (Supplementary Fig. 15 ), and these theoretical values are then connected with experimentally measured overpotentials. The ML model achieves remarkable accuracy in predicting true overpotentials (Fig. 2d ).

figure 2

a , Representative structure of multimetallic hydroxides generated by classical MD simulation. b , Reaction mechanism of the OER. c , Statistics of three OER descriptors, ∆ G OH* , ∆ G O*−OH* and ∆ q , predicted by the NN. d , The prediction results of measured OER overpotentials by predictive model that was calibrated by experiments. In c and d , r is the Pearson correlation coefficient.

Source data

High-throughput automated synthesis-characterization-performance optimization executed by the ai chemist.

Using the LIBS-determined elemental composition of each Martian ore in Fig. 3a (here we use Martian meteorites to represent in situ Martian ores), the AI chemist prepared 243 different formulas with randomly selected compositions of six metal elements, performed electrocatalytic OER testing using each of them as catalyst and measured overpotentials by analysing linear sweep voltammetry (LSV) polarization curves at a current density of 10 mA cm −2 ( η 10 ) per geometric area. The reason for choosing this specific current density is that it is approximately the current density expected at the anode of a 23% efficient solar-to-fuels conversion device under 1-sun illumination received on Mars 18 , 19 . This preliminary screening generated an array of η 10 values ranging from 482.2 to 1,056.2 mV (Fig. 3b ). Then we trained the second NN model by using three computed OER activity descriptors and 243 sets of compositions as inputs and their corresponding experimental overpotentials as output (Fig. 2d ). By concatenating these two NN models, the OER overpotentials for all 29,902 compositions can be easily predicted, creating a much larger dataset for Bayesian optimization 10 to generate the optimal formula for a desired OER catalyst (Fig. 3c ).

figure 3

a , Representative LIBS spectral curves of five meteorite specimens in 200–380 nm with the ownership of main elemental emission lines. b , The η 10 values of 243 ‘trial-and-error’ experiments performed by the mobile robot and workstations. The blue line indicates the best sample, ‘Experiment No. 197’ (in c and d , referred to as Exp-197), with the lowest overpotential among 243 pilot experiments; green line represents the catalyst obtained through experimental-data-guided local optimal search (in c and d , referred to as Exp-guided OPT); pink line demonstrates the result from Bayesian optimization using both simulated and experimental datasets (in c and d , referred to as Model-guided OPT). The pink star highlights the lowest OER overpotential, suggesting that the ML model using both theoretical and experimental data achieves a global best synthetic formula, outperforming any other approach. c , LSV curves collected at a sweep rate of 5 mV s −1 in 1 M KOH electrolyte. Insets are photographs of catalyst materials synthesized with components based on Martian meteorite composition, from left to right, Exp-197, Exp-guided OPT and Model-guided OPT. d , Kiviat diagram of elemental ratios. e , LSV curves of Model-guided OPT collected in CO 2 -saturated 2.8 M Mg(ClO 4 ) 2 electrolyte at 23 °C and −37 °C, respectively. f , Corresponding Tafel plots for the anodic water oxidation derived from the LSV data in e to evaluate the reaction kinetics.

As the Kiviat plot indicates (Fig. 3d ), the optimal compositions identified by the Bayesian model differ greatly from those of the best sample, namely Experiment No. 197 (Exp-197) from the pilot experiments, indicating that Bayesian optimization based on both simulated and experimental datasets can surpass experimental-data-guided local search. Meanwhile, the product with optimal composition predicted by the model that uses theoretical data also gives worse performance than the one from Bayesian optimization that relies on both simulated and experimental data (Supplementary Fig. 16 ). The catalyst with optimal composition (Model-guided OPT) was synthesized and verified by the AI chemist to have η 10  = 445.1 mV, showing a substantial improvement (37.1 mV lower) with respect to the best result from a purely experimental search. Bayesian optimization suggested a metal composition that was almost identical to that suggested by the grid point scanning in the simulation with the re-trained NN model, but with much less time consumption, suggesting that Bayesian optimization is more effective in finding a solution (Supplementary Table 3 ). For comparison, we also made a test using the experimental data only as input for Bayesian optimization. The resultant optimal composition of meteorites (Exp-guided OPT) gives η 10  = 467.4 mV, which is very close to the best result among the 243 pilot experiments (Fig. 3d ). Hence, the intrinsic limitation of local optimization with limited experimental data would likely be overcome by concatenating NN models trained from both theoretical and experimental data. After all, attempting to achieve the global best synthetic formula by exhaustive trial-and-error approach requires 3,764,376 possible experimental traversal searches (Supplementary Note 1 ), which is a nearly impossible task. The synthetic formulas of the several studied catalysts are listed in Supplementary Table 4 , clearly and quantitatively comparing the differences in metal ratios among them. We also synthesized catalysts using only one meteorite as feedstock and found that all performances were inferior to the local optimum solution found (Exp-197) (Supplementary Fig. 17 ).

After determining the optimal metal ratio with minimum overpotential, we performed detailed comparisons of other electrochemical parameters (Supplementary Figs. 18 – 20 ). We derived the trend of the reaction activation energy from the Tafel slope. The Model-guided OPT in 1.0 M KOH required a low value of only 61.35 mV dec −1 to reach η 10 , outperforming Exp-197 (83.59 mV dec −1 ) and Exp-guided OPT (65.02 mV dec −1 ) catalysts, indicating that it possesses a favourable kinetic process for an OER. The electrochemical surface area reflects the chemisorption capacity of the reacting substrate and the exposure of surface active sites. This parameter can be estimated by measuring the Helmholtz double-layer capacitance, allowing comparison of the intrinsic electrochemical activity of different catalysts 20 . We found that the Model-guided OPT catalyst possesses the highest double-layer capacitance, which is about twice that of either the Exp-197 or Exp-guided OPT catalyst. This result implies that H 2 O molecules can be in close contact with the surface of this catalyst and that both the input of electrolyte and diffusion of the gaseous O 2 product are effectively accelerated. Electrochemical impedance spectroscopy is a technique that probes the internal processes of an electrochemical system and allows measurement of the operating state of the electrodes for kinetics study. By measuring the location and size of the semicircular region in the Nyquist plots, the voltages of solution and working electrode resistance losses can be obtained, which helps to analyse impedance changes as an aid in the assessment of electrocatalytic efficiency. We found that the semicircular diameters of the Exp-197, Exp-guided OPT and Model-guided OPT catalysts decrease sequentially, indicating that their charge transfer (or Faradaic) resistances follow the same trend.

Feasibility validation of oxygen production under simulated Martian environment

To verify the usability of catalysts under the low-temperature condition on the Martian surface, we performed an experiment based on the previous work of Gayen et al. 1 and the fact that large-scale water resources have been found in the Martian regolith 21 , 22 . The operating conditions on the Martian surface were constructed with a 2.8 M Mg(ClO 4 ) 2 brine solution as the electrolyte, platinum mesh as counter electrode and robust Ag wire as the reference at 23 °C and −37 °C (Fig. 3e ). The LSV polarization curves suggest low voltage values of 1.5685 V and 1.7289 V to reach the current density of 10 mA cm −2 with Tafel slopes of 174.1 mV dec −1 and 200.3 mV dec −1 , respectively (Fig. 3f and Supplementary Fig. 21 ).

Long-term stability is crucially important for the practical application of OER catalysts. We performed cycling stability tests on the catalyst under various conditions by applying a certain voltage to the catalyst working electrode and assembling the electrolyser to maintain its initial oxygen production current at 10 mA cm −2 in a classical three-electrode system (Fig. 4a–f ). Prolonged testing showed that the as-prepared Model-guided OPT catalyst works steadily at the current density of 10 mA cm −2 for more than 550,000 s (~153 h) in 1 M KOH at 23 °C and 350,000 s (~97 h) in 2.8 M Mg(ClO 4 ) 2 brine solution at −37 °C (Fig. 4g ), indicating that this AI-chemist-designed catalyst is as stable as other state-of-the-art OER catalysts. It can also be estimated that in 1 M KOH, the catalyst made by AI chemist can achieve an average O 2 production rate of 59.08 g h −1  m −2 . For a Martian station room with 300 m 3 volume (100 m 2 area and 3 m height) coated with the produced OER catalyst film on its roof, it will take about 15.2 h to achieve oxygen self-sufficiency. This process could be accelerated with the catalyst directly grown on the conductive nickel foam substrate as it is synthesized, which maintains considerably efficient and stable O 2 production capability at an even higher current density condition (Supplementary Fig. 22 ), albeit requiring larger area solar panels to generate more electricity to boost the OER reaction. It is likely that given more types of metal elements in Martian ores and advanced mineral refining facilities, the performance of Mars-mineral-derived chemicals and materials can be further improved in the future.

figure 4

a , Positioning of the electrochemical workstation. b , Clamping of the carbon paper. c , Assembly of electrodes. d , Dripping of the catalyst ink. e , Placement of the working electrodes into the electrolyser. f , Sunlight-driven OER process. g , The time-dependent current density curve of a Mode-guided OPT sample on carbon-paper substrate acquired at working current density of 10 mA cm −2 under 1 M KOH alkaline conditions at 23 °C and CO 2 -saturated 2.8 M Mg(ClO 4 ) 2 electrolyte at −37 °C, respectively.

Our study provides a demonstration that an advanced AI chemist can, without human intervention, synthesize OER catalysts on Mars from local ores. This system has demonstrated its ability to perform all required experimental steps, including raw material analysis, pretreatment, synthesis, characterization and performance testing with high precision and also shown its intelligent analysis power in identifying the best formula for a Martian OER catalyst from millions of possible combinations. Particularly powerful is the in situ optimization, which seamlessly combines the experimental data and computational data during the synthesis process, greatly accelerating the generation of a reliable model and finding of an optimal formula. The established protocol and system, which are generic and adaptive, are expected to advance automated material discovery and synthesis of chemicals for the occupation and exploration of extraterrestrial planets.

Chemicals and materials

NaOH (99.9%), KOH (99.9%), HCl (37% trace metals), K 3 [Fe(CN) 6 ] (99%), K 4 [Fe(CN) 6 ]•3H 2 O (99.5%), anhydrous ethanol and 5 wt% Nafion 117 solution were purchased from Sigma-Aldrich. The counter electrode of graphite rod and reference electrodes of Ag/AgCl in saturated KCl were purchased from CH Instruments. The deionized water (18.2 MΩ cm −1 ) used for the feedstocks solution and aqueous electrolytes preparation were made with a Milli-Q EQ 7000 Ultrapure water purification system.

Catalyst synthesis

Five different categories of meteorites that come from or have been confirmed to exist on Mars were selected 23 , 24 ; complete information describing these approved meteorites can be found on the website of the Meteoritical Bulletin Database available at . We digested various masses of these individual meteorites in 1 mol l −1 hydrochloric acid solution based on the results of elemental analysis by a LIBS spectrometer, which was used to configure the feedstocks solution to control the total mass concentration of the six key catalytic metals (that is, Fe, Mn, Ni, Ca, Mg and Al) to 200 mg l −1 . For the catalyst preparation, the AI chemist set the addition amount of total feedstocks solution to reaction vial to 10 g at the liquid-dispensing workstation, but randomly varied the proportion of each feedstock entered. In this manner, the atomic ratio of metals in the final product can be finely adjusted. Afterwards, 3 g of aqueous NaOH solution with a concentration of 4 mol l −1 was added to the reaction vials and stirred for 5 minutes, followed by centrifugation at 7,500 g for 5 minutes, aspiration of the upper waste solution, washing with anhydrous ethanol and drying at 60 °C. The described synthetic procedure was done to perform the initial search for the optimal metal ratio in a catalyst and was performed for 243 groups of experiments. During the whole process, the intelligent ‘brain’ of the AI-chemist system automatically generates .xml execution files and sends them to the experimental robot, which—with various synthesis and testing workstations—sequentially automates the preparation of the catalyst material. Similarly, the synthesis of samples Exp-guided OPT and Model-guided OPT is based on the same method as above, except that the respective metal ratios are determined and given by the intelligence ‘brain’ and converted by a transformation matrix.

Analysis of metal content in meteorites by LIBS

LIBS is a rapid chemical analysis technology that offers many compelling advantages compared to other elemental analysis techniques in geoscience. The main physical process that constitutes the essence of LIBS technology is the formation of a high-temperature plasma created by ultrafast laser pulses (UFLPs). When the UFLP beam is focused on the surface, a small portion of sample mass is ablated, a process called laser ablation. This ablated mass further interacts with the trailing portion of the UFLP to form a short-lived high-temperature plasma containing free electrons, excited atoms and ions. When the laser pulse is terminated, the plasma cools and the electrons in atoms or ions at the excited state decay to their natural ground state. Correspondingly, the wavelengths of the emitted photons are inversely proportional to the energy difference between the excited and ground states, so that each element has its own set of characteristic emission wavelengths, a fingerprint signature, which is then collected and coupled to the spectrometer detector module for LIBS spectroscopy. Each element in the periodic table is associated with a unique LIBS spectral peak. The high energy density of the focused UFLP allows the excitation of material in any physical state (in our case, solid) to form a plasma, allowing the LIBS technique to analyse samples and assess the relative abundance of each constituent element.

Our LIBS workstation consists of a researcher-developed sample feed system, a nanosecond laser generator (Quantel Viron), a fibre-optic spectrometer (AvaSpec-ULS2048CL-2-EVO), an optical system and a high-performance computer. The feed system consists of a motorized delivery track, a slowly rotating sample stage and a motor control unit. The control unit is designed independently, with the main control chip being the Atmel-produced MEGA 2560V chip. The robotic arm of the AI chemist places the sample on the stage, which is then delivered by the feed system for laser irradiation with slow rotation so that different points on the surface are excited by the laser to obtain an unattenuated signal. The pulsed laser emitted from the laser generator is focused on the sample surface by the optical system to produce a transient high-temperature plasma. The signal is captured by the optical fibre of the spectrometer. A researcher-written programme in the computer controls the measurement process automatically and acquires spectral data from the spectrometer for subsequent analysis and processing.

The LIBS spectra were collected under optimized conditions: laser pulse energy, 105 mJ; spot diameter, 2 mm; spectrometer slit width, 15 mm; gate width, 1 ms and acquisition delay, 180 ms. In total, 388 sets of data points were obtained, these spectra were accumulated, and then the peak line region of the target element was marked according to the elemental peak lines obtained from the NIST database ( ) and the baseline correction was performed for peak line region. After wavelet filtering, the best Lorentz peak shape and the offset of the actual peak line relative to its standard spectrum were obtained by fitting with the Levenberg–Marquadt method. Here, the half-height width of the peak, the wavelength of crest and the signal intensity were used as intrinsic characteristics. The preliminary element content was calculated by linear regression. Subsequently, the top 50 data points in the original spectrum with the highest correlation with the elemental content (found by the LASSO algorithm after normalization from the pretraining set data) are transformed in the same way as the pretraining set data and then entered into a pretrained backpropagation NN together with the preliminary content calculated in the previous step to obtain a more accurate elemental content. Similarly, the above analysis is repeated for each targeted element to obtain the exact content with a relative error within ±5%.

Transfer metals molar ratio to Martian ores mass ratio

The feedstock solutions are prepared by the following procedure: Take out 271.05 mg of Aletai, 567 mg of NWA 8171, 563.2 mg of NWA 13669, 935 mg of NWA 12564, 688.5 mg of Hassi Messaoud 001, each dissolved in 1 l of acidic solution to prepare the feedstock solution for the experiments. In this way, the total mass concentrations of metal ions in all feedstock solutions are controlled at about 200 mg l −1 ; the concentrations of metals are listed in Supplementary Table 1 .

Because the ML-model-predicted results are the metal molar ratios, we prepared a researcher-developed software programme to transfer metals molar ratio to Martian ores mass ratio for convenient robotic weighting operation. The software is developed using Python, and it is also converted to a windows-based executable programme (Supplementary Fig. 15 ). The source code is as follows:

 import tkinter as tk

 import numpy as np

 from scipy.linalg import solve

 window = tk.Tk()

 window.title(‘Transfer ratio’)


 s1 = tk.Label(window,text = ‘Fe’)

 a1 = tk.Entry(window,show=None)

 s2 = tk.Label(window,text = ‘Mn’)

 a2 = tk.Entry(window,show=None)

 s3 = tk.Label(window,text = ‘Ni’)

 a3 = tk.Entry(window,show=None)

 s4 = tk.Label(window,text = ‘Ca’)

 a4 = tk.Entry(window,show=None)

 s5 = tk.Label(window,text = ‘Mg’)

 a5 = tk.Entry(window,show=None)

 s6 = tk.Label(window,text = ‘Al’)

 a6 = tk.Entry(window,show=None)

 def transfer():

  b1 = a1.get()

  b2 = a2.get()

  b3 = a3.get()

  b4 = a4.get()

  b5 = a5.get()

  b6 = a6.get()

metal_ratio = np.array([float(b1),float(b2),float(b3),float(b4),float(b5),float(b6)])

abundance = np.array([[3.303,0,0.328,0.0055,0.00155,0.00149],





transfer_matrix = abundance.T

meteorite_ratio = solve(transfer_matrix[0:5], metal_ratio[0:5])

meteorite_ratio = meteorite_ratio/sum(meteorite_ratio)

abcdelist = [0.27105,0.567,0.5632,0.935,0.6885]

meteorite_ratio = meteorite_ratio*10*abcdelist

meteorite_ratio = meteorite_ratio.round(4)

result = ‘,’.join(str(i) for i in meteorite_ratio)

t.insert(‘insert’,‘meteorite ratio (A,B,C,D,E) is: ‘+result + ‘\n’)

button = tk.Button(window,

text = ‘Transfer’,

width=15, Height=2,



t = tk.Text(window)


OER measurement under 1 M KOH alkaline condition

All the electrochemical measurements were conducted at the electrochemical workstation (CHI660E, CH Instruments) in a standard three-electrode setup with the catalyst derived from Martian meteorites as the working electrode, a graphite rod as the counter electrode and Ag/AgCl in saturated KCl as reference electrode. All the electrocatalytic OER performance was studied under alkaline conditions (1 mol l −1 KOH). The applied potential was calibrated to reversible hydrogen electrode (RHE) following the equation E RHE  =  E Ag/AgCl  + 0.0591 × pH + 0.197 V. Unless otherwise specified, neither iR-compensation nor background current correction was applied. For the working electrode preparation, the as-prepared catalyst dispersing in 5 ml of a mixed solution of ethanol (4.8 ml) and 5 wt% Nafion (0.2 ml) under magnetic stirring to form a uniform catalyst ink. Then, 200 μl of the resulting catalyst ink was drop-casted onto a carbon paper with the loading area of 2.5 × 2 cm 2 , and the corresponding final metal loading was calculated to be 0.08 mg cm −2 . Cyclic voltammetry activation curve was performed 40 times from 1.0 V to 1.5 V with respect to the RHE reference at a sweep rate of 50 mV s −1 . LSV measurements were performed from 1.0 V to 2.0 V with respect to the RHE reference at a scan rate of 5 mV s −1 . Tafel slope ( b ) is obtained by fitting the linear portion according to the Tafel equation ( η  =  a  + blog j ) using the overpotential ( η ) as a function of the logarithmic scale of current density (log j ). Electrochemical impedance spectroscopy measurements performed at an overpotential of 0.4 V for working electrodes. Electrochemical active surface areas are evaluated based on the double-layer capacitance via the analysis of a series of cyclic voltammetry measurements performed within the non-Faradaic potential region (1.05 to 1.15 V with respect to the RHE reference) at various scan rates (10, 20, 40, 60, 80, 100, 120, 140, 160, 180 and 200 mV s −1 ). The chronoamperometry ( i – t ) test was collected at a constant potential at 1.7 V with respect to the RHE reference for 550,000 seconds. All the electrochemical characterizations can be performed automatically by one-click measurements and generate the experimental reports using a researcher-written Python code. To grow the catalyst on nickel foam substrate for oxygen production at industrial current density, we added feedstocks solution prepared from five Martian meteorites to the autoclave reactor, followed by NaOH addition to adjust the pH to 5–6, then 0.2 g of urea was added to dissolve, and finally cleaned nickel foam with thickness of 2 mm was placed vertically, encapsulated and held at 130 °C for 10 hours. When the reaction is completed, the nickel foam is taken out, washed and dried for OER performance testing in compliance with the described three-electrode system.

In situ electrocatalytic oxygen generation experiments under simulated Martian surface environmental conditions

As the Martian surface is well below 0 °C for most of the Mars year and its atmosphere is rich in CO 2 , we used aqueous Mg(ClO 4 ) 2 solution (pH ≈ 7) with a concentration of 2.8 M as a mimic of the brine solution already explored on Mars and then used a dry ice solution of ethanol-ethylene glycol mixture at a constant temperature of −37 °C for OER testing 25 . Considering such a low temperature, the conventional Ag/AgCl electrode is no longer suitable as a reference; therefore, we used Ag wire (99.99%, φ  = 1.5 mm) as a reference electrode at low-temperature conditions and potassium ferrocyanide-potassium ferricyanide oxidation-reduction potential buffer as internal standard to determine the potential of Ag wire at approximately 0.427 V with respect to an RHE reference. All electrochemical test steps and data processing procedures are similar to those performed in 1 M KOH, except that the corresponding voltage window is changed and the Mg(ClO 4 ) 2 solution is saturated with CO 2 (99.999%) prior to testing.

Theoretical calculations

Md simulations.

To extract structural features of high-entropy hydroxides, we sampled one million equilibrated structures for each of 29,902 unique formulas of six-metallic elements (Fe, Ni, Mn, Ca, Mg and Al) using classical MD simulation. The initial configuration of each composition was generated by randomly placing 60 different metal cations and sthe corresponding number of hydroxyl anions for maintaining neutrality into a cubic box of 3 × 3 × 3 nm 3 using GROMACS 26 . The universal force field 27 was adopted and all parameters for high-entropy hydroxides were generated by the LAMMPS Interface programme 28 . The cutoff distances for both Lennard–Jones and Coulombic potential were set to be 12.5 Å. Then, each initial structure was pre-equilibrated by energy minimization. In a production MD run, a trajectory of 1 ns with a time step of 1 fs was collected in an NPT ensemble with P  = 1 atm and T  = 2,000 K using the Nosé–Hoover barostat and thermostat 29 , 30 . For each trajectory, we retrieved 100 configurations in an evenly divided 10 ps interval and computed averaged metal–metal and metal–oxygen distances as structure features of these high-entropy hydroxides. All force field base simulations were carried out with the LAMMPS package 31 .

DFT calculation

To describe the OER activity of high-entropy hydroxides, DFT calculations on the simplified bimetallic hydroxide model with the information of statistical structure features of each unique composition embedded were performed using the Perdew–Burke–Ernzerhof functional 32 and the projector augmented wave method 33 as implemented in the Vienna ab initio simulation package 34 . The kinetic energy cutoff of the plane-waved basis set was 400 eV. The Brillouin zone was sampled with 3 × 2 × 1 Monkhorst–Pack k-mesh with the vacuum size chosen to be 15 Å to avoid interaction between two layers for all structures. The long-range van der Waals interaction corrections were described using Grimme’s D3 correction 35 . All geometry but the metal–( η 2 -oxygen) 2 –metal moiety is allowed to relax. The convergences of total energy for wave function self-consistency and force between atoms for optimization were set to be 10 −5  eV and 0.01 eV Å −1 , respectively.

Calculation for the free energies

The Gibbs free energies ∆ G OH* and ∆ G O*−OH*  = ∆ G O* −∆ G OH* and ∆ q (the amount of charge transferred for hydroxyl adsorption on the activate site) were used as computational descriptors of OER activity. For all possible combinations of dual metal atoms in every high-entropy hydroxide, their OER descriptors were calculated using the following procedure.

The elementary steps of hydroxyl adsorption and oxygen adsorption can be given as:

Under zero potential, the Gibbs free energy of each elementary step was given by the expression:

where ∆ E is the change in reaction energy. The ∆ZPE is the zero-point energy change, and ∆ S is the entropy change for each elementary step with the temperature at 298.15 K.

NN architecture

The first NN model—which uses information of metal composition as input and the DFT calculated three catalytic properties (∆ G OH* , ∆ G O*−OH* , ∆ q ) as output—comprises one input layer, two hidden layers and one output layer. The number of neurons in both hidden layers is 512. To link computed catalytic properties and experimentally measured overpotential, the second NN model was built with one input layer, three hidden layers involving 128 neurons each and one output layer. For the training of each NN, the dataset was divided into two subsets, one for training (80%) and the other for testing (20%). The NN model was trained with a backpropagation algorithm and the Rectified Linear Unit activation function 36 as implemented in TensorFlow 37 .

Two NNs were combined to create a predicting model that used the metal composition as a descriptor to estimate the real overpotential. A Bayesian approach, taking above predicting model as objective function, was then applied to identify the optimal metal composition with the highest catalytic activity. The Bayesian optimization loop consisted of 280 iterations, and the surrogate model was a basic Gaussian process, which could capture the uncertainty and noise in the data and handle different types of objective functions. We also used the upper confidence bound as our acquisition function, which balanced exploration and exploitation by adding a positive term depending on the standard deviation to the mean estimate of the objective function.

Data availability

The data that support the findings of this study are available in the paper, its Supplementary Information and Supplementary Video 1 .

Code availability

The code used for training an NN model for OER prediction with theoretical data and robot-driven experimental data is available on GitHub at .

Gayen, P., Sankarasubramanian, S. & Ramani, V. K. Fuel and oxygen harvesting from Martian regolithic brine. Proc. Natl Acad. Sci. USA 117 , 31685–31689 (2020).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Hoffman, J. A. et al. Mars Oxygen ISRU Experiment (MOXIE)—preparing for human Mars exploration. Sci. Adv. 8 , eabp8636 (2022).

Wade, J., Dyck, B., Palin, R. M., Moore, J. D. P. & Smye, A. J. The divergent fates of primitive hydrospheric water on Earth and Mars. Nature 552 , 391–394 (2017).

Article   CAS   PubMed   Google Scholar  

Orosei, R. et al. Radar evidence of subglacial liquid water on Mars. Science 361 , 490–493 (2018).

Kruyer, N. S., Realff, M. J., Sun, W., Genzale, C. L. & Peralta-Yahya, P. Designing the bioproduction of Martian rocket propellant via a biotechnology-enabled in situ resource utilization strategy. Nat. Commun. 12 , 6166 (2021).

Yao, Y. et al. Extraterrestrial photosynthesis by Chang’E-5 lunar soil. Joule 6 , 1008–1014 (2022).

Article   CAS   Google Scholar  

Burger, B. et al. A mobile robotic chemist. Nature 583 , 237–241 (2020).

Rohrbach, S. et al. Digitization and validation of a chemical synthesis literature database in the ChemPU. Science 377 , 172–180 (2022).

Zhu, Q. et al. An all-round AI-chemist with a scientific mind. Natl Sci. Rev. 9 , nwac190 (2022).

Pyzer-Knapp, E. O., Chen, L., Day, G. M. & Cooper, A. I. Accelerating computational discovery of porous solids through improved navigation of energy-structure-function maps. Sci. Adv. 7 , eabi4763 (2022).

Article   Google Scholar  

Butler, K. T., Davies, D. W., Cartwright, H., Isayev, O. & Walsh, A. Machine learning for molecular and materials science. Nature 559 , 547–555 (2018).

Sanchez-Lengeling, B. & Aspuru-Guzik, A. Inverse molecular design using machine learning: generative models for matter engineering. Science 361 , 360–365 (2018).

Bai, L., Hsu, C.-S., Alexander, D. T. L., Chen, H. M. & Hu, X. Double-atom catalysts as a molecular platform for heterogeneous oxygen evolution electrocatalysis. Nat. Energy 6 , 1054–1066 (2021).

Lin, C. et al. In-situ reconstructed Ru atom array on α-MnO 2 with enhanced performance for acidic water oxidation. Nat. Catal. 4 , 1012–1023 (2021).

Xu, H., Cheng, D., Cao, D. & Zeng, X. C. A universal principle for a rational design of single-atom electrocatalysts. Nat. Catal. 1 , 339–348 (2018).

Craig, M. J. et al. Universal scaling relations for the rational design of molecular water oxidation catalysts with near-zero overpotential. Nat. Commun. 10 , 4993 (2019).

Article   PubMed   PubMed Central   Google Scholar  

Tao, L. et al. Charge transfer modulated activity of carbon-based electrocatalysts. Adv. Energy Mater. 10 , 1901227 (2020).

Walter, M. G. et al. Solar water splitting cells. Chem. Rev. 110 , 6446–6473 (2010).

Appelbaum, J. & Flood, D. J. Solar radiation on Mars. Sol. Energy 45 , 353–363 (1990).

McCrory, C. C. L., Jung, S., Peters, J. C. & Jaramillo, T. F. Benchmarking heterogeneous electrocatalysts for the oxygen evolution reaction. J. Am. Chem. Soc. 135 , 16977–16987 (2013).

Hecht, M. H. et al. Detection of perchlorate and the soluble chemistry of Martian soil at the Phoenix lander site. Science 325 , 64–67 (2009).

Cull, S. C. et al. Concentrated perchlorate at the Mars Phoenix landing site: evidence for thin film liquid water on Mars. Geophys. Res. Lett. 37 , L22203 (2010).

Schröder, C. et al. Meteorites on Mars observed with the Mars exploration rovers. J. Geophys. Res. 113 , E06S22 (2008).

Google Scholar  

Ashley, J. CosmoELEMENTS: the study of exogenic rocks on Mars—an evolving subdiscipline in meteoritics. Elements 11 , 10–11 (2015).

Jensen, C. M. & Lee, D. W. Dry-ice bath based on ethylene glycol mixtures. J. Chem. Educ. 77 , 629 (2000).

Abraham, M. J. et al. GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1–2 , 19–25 (2015).

Mayo, S. L., Olafson, B. D. & Goddard, W. A. DREIDING: a generic force field for molecular simulations. J. Phys. Chem. 94 , 8897–8909 (1990).

Boyd, P. G., Moosavi, S. M., Witman, M. & Smit, B. Force-field prediction of materials properties in metal-organic frameworks. J. Phys. Chem. Lett. 8 , 357–363 (2017).

Nosé, S. A unified formulation of the constant temperature molecular dynamics methods. J. Chem. Phys. 81 , 511–519 (1984).

Hoover, W. G. Canonical dynamics: equilibrium phase-space distributions. Phys Rev A (Coll Park) 31 , 1695–1697 (1985).

Plimpton, S. Fast parallel algorithms for short-range molecular dynamics. J. Comput. Phys. 117 , 1–19 (1995).

Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett 77 , 3865–3868 (1996).

Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50 , 17953–17979 (1994).

Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6 , 15–50 (1996).

Grimme, S., Ehrlich, S. & Goerigk, L. Effect of the damping function in dispersion corrected density functional theory. J Comput. Chem. 32 , 1456–1465 (2011).

Maas, A. L., Hannun, A. Y. & Ng, A. Y. Rectifier nonlinearities improve neural network acoustic models. In Proc. 30th International Conference on Machine Learning (ICML) (eds Dasgupta, S. & McAllester, D.) 3 (ACM Press, 2013).

Abadi, M. et al. TensorFlow: a system for large-scale machine learning. In Proc 12th USENIX Conference on Operating Systems Design and Implementation (eds Keeton, K. & Roscoe, T.) 265–283 (USENIX Association, 2016).

Download references


Y.L. acknowledges funding support for this research from the Innovation Program for Quantum Science and Technology (Grant 2021ZD0303303). J.J. gratefully acknowledges financial support by the National Natural Science Foundation of China (Grants 22025304, 22033007) and the CAS Project for Young Scientists in Basic Research (Grant YSBR-005). Q.Z. gratefully acknowledges the financial support of the National Natural Science Foundation of China (Grant 22103076) and Anhui Provincial Natural Science Foundation (Grant 2108085QB63). We also gratefully acknowledge the USTC Center for Micro- and Nanoscale Research and Fabrication for providing experimental resources and the USTC supercomputing centre for providing computational resources.

Author information

These authors contributed equally: Qing Zhu, Yan Huang, Donglai Zhou, Luyuan Zhao.

Authors and Affiliations

Key Laboratory of Precision and Intelligent Chemistry, Hefei National Research Center for Physical Sciences at the Microscale, School of Chemistry and Materials Science, University of Science and Technology of China, Hefei, China

Qing Zhu, Yan Huang, Donglai Zhou, Luyuan Zhao, Lulu Guo, Ruyu Yang, Zixu Sun, Man Luo, Hengyu Xiao, Baicheng Zhang, Jiaqi Cao, Guozhen Zhang, Song Wang, Huirong Li, Jun Jiang & Yi Luo

School of Information Science and Technology, University of Science and Technology of China, Hefei, China

Fei Zhang, Xinsheng Tang, Xuchun Zhang, Tao Song, Xiang Li, Baochen Chong, Junyi Zhou, Yihan Zhang, Shuang Cong & Weiwei Shang

Hefei JiShu Quantum Technology Co. Ltd., Hefei, China

Guilin Ye & Wanjun Zhang

Materials Interfaces Center, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China

Haitao Zhao

Deep Space Exploration Laboratory, Hefei, China

Li-Li Ling & Zhe Zhang

School of Aerospace Engineering, Beijing Institute of Technology, Beijing, China

Hefei National Laboratory, University of Science and Technology of China, Hefei, China

Jun Jiang & Yi Luo

You can also search for this author in PubMed   Google Scholar


These authors contributed equally: Q.Z., Y.H., D.Z., L.Z. Q.Z. planned and conducted the robotic experiments and collected and analysed the experiment data. Y.H., D.Z., L.Z. and H.L. performed theoretical simulations and ML training. L.G., R.Y., Z.S. and M.L. assisted with the spectroscopic characterization and data analysis. H.X., B.Z. and J.C. were responsible for writing test scripts. X.T. and Y.Z. contributed to the development of robotic operation module, robotic arm motion planning and force control. J.Z. and B.C. helped with the robot platform communication, SLAM, platform motion planning and navigation. T.S. planned robot movement and operation task management system. X.L. and S.C. managed the scheduling optimization of robot experimental tasks at various workstations. X.Z. developed the robotic visual localization algorithm. F.Z. and W.S. designed the entire robot system. G.Y. and W.Z. worked on non-standardized equipment development. S.W., G.Z. and H.Z. contributed to the original draft preparation. L.-L.L. and Z.Z. assisted in the design and execution of experiments under simulated Martian environments. J.J. and Y.L. conceptualized the study, developed the methodology and conducted the investigation and wrote, reviewed and edited the paper. All authors participated in discussions and revisions and provided comments on the paper.

Corresponding authors

Correspondence to Weiwei Shang , Jun Jiang or Yi Luo .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Synthesis thanks Leroy Cronin, Zhigang Zou and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Peter Seavill, in collaboration with the Nature Synthesis team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information.

Experimental details, Supplementary Figs. 1–22, Note 1 and Tables 1–4.

Supplementary Video 1

This video showcases the capabilities of the AI chemist in synthesizing and optimizing oxygen-producing catalysts from Martian meteorites. The process involves automated analysis of Martian ore, catalyst synthesis, characterization, intelligent computing and OER performance testing, which highlights the integration of robotics and AI for complex materials design and manufacture under challenging circumstances.

Source Data Fig. 2

Source data for Fig. 2.

Source Data Fig. 3

Source data for Fig. 3.

Source Data Fig. 4

Source data for Fig. 4.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and Permissions

About this article

Cite this article.

Zhu, Q., Huang, Y., Zhou, D. et al. Automated synthesis of oxygen-producing catalysts from Martian meteorites by a robotic AI chemist. Nat. Synth (2023).

Download citation

Received : 11 March 2023

Accepted : 27 September 2023

Published : 13 November 2023


Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

This ai robot chemist could make oxygen on mars.

  • Jonathan O'Callaghan

Nature (2023)

Quick links

  • Explore articles by subject
  • Guide to authors
  • Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

research paper machine learning

--> AGU