A Survey on Machine Learning-Based HPC I/O Analysis and Optimization

As one of the earth's key ecosystems, rivers have been intensively studied and modelled through the application of machine learning (ML). With the amount of large data available, these computer algorithms are ever increasing in numerous fields, although there is ongoing scepticism and scholars still question the actual impact and deliverables of algorithms. This study aims to provide a systematic review of the state‐of‐the‐art ML‐based techniques, trends, opportunities and challenges in river research by applying text mining and automated content analysis. Unsupervised and supervised learning have dominated river research while neural networks and deep learning have also gradually gained popularity. Matrix factorisation and linear models have been the most popular ML algorithms, with around 1300 and 800 publications on these topics in 2020 respectively. In contrast, river researchers have had few applications in multiclass and multilabel algorithm, associate rule and Naïve Bayes. The current article proposes an end‐to‐end workflow of ML applications in river research in order to tackle major ML challenges, including four steps: (1) data collection and preparation; (2) model evaluation and selection; (3) model application; and (4) feedback loops. Within this workflow, river modellers have to balance numerous trade‐offs related to model traits, such as complexity, accuracy, interpretability, bias, data privacy and accessibility and spatial and temporal scales. Any choices made when balancing the trade‐offs can lead to different model outcomes affecting the final applications. Hence, it is necessary to carefully consider and specify modelling goals, understand the data collected and maintain feedback loops in order to continuously improve model performance and eventually reach the research objectives. Moreover, it remains crucial to address the users' needs and demands that often entail additional elements, such as computational cost, development time and the quantity, quality and compatibility of data. Furthermore, river researchers should account for new technologies and regulations in data collection and protection that are transforming the development and applications of ML, most notably data warehouse and information management with multiple‐cycles that are becoming a cornerstone of the integration of ML in decision‐making in river and ecosystem management.

BackgroundThe exponential growth of publications regarding the application of machine learning (ML) tools in medicine highlights the significant potential for ML to revolutionize the field. Despite the multitude of literature surrounding this topic, there are limited publications addressing the implementation and feasibility of ML models in clinical practice. Currently, Machine Learning Operations (MLOps), a set of practices designed to deploy and maintain ML models in production, is used in various information technology and industrial settings. However, the MLOps pipeline is not well researched in medical settings, where multiple barriers exist to implementing ML pipelines into practice.ObjectiveThis study aims to detail how MLOps is implemented in health care and propose a maturity framework for the health care implementations.MethodsA scoping review search was conducted according to the Joanna Briggs Institute Manual for Evidence Synthesis. Results were synthesized using the 3-stage basic qualitative content analysis. We searched 4 databases (eg, MEDLINE, Embase, Web of Science, and Scopus) to include any studies that involved proof of concept or real-world implementation of MLOps in health care. Studies not reported in English were excluded.ResultsA total of 19 studies were included in this scoping review. The MLOps workflow identified within the studies included (1) data extraction (19/19 studies), (2) data preparation and engineering (18/19 studies), (3) model training (19/19 studies), (4) measured ML metrics and model evaluation (17/19 studies), (5) model validation and test in production (14/19 studies), (6) model serving and deployment (15/19 studies), (7) continuous monitoring (14/19 studies), and (8) continual learning (13/19 studies). We proposed a 3-stage MLOps maturity framework for health care based on existing studies in the field, that is, low (5/19 studies), partial (1/19 studies), and full maturity (13/19 studies). There were 8/19 studies that discussed ethical, legislative, and stakeholder considerations for MLOps implementations in health care settings.ConclusionsWe investigated the implementation of MLOps in health care with a corresponding maturity framework. It is evident that only a limited number of studies reported the implementation of ML in health care contexts. Hence, it is imperative that we shift our focus toward creating an environment that supports the development of ML health care applications, such as improving existing data infrastructure, and engaging partners to support the development of MLOps applications. Specifically, we can include patients, policymakers, and health care professionals in the creation and implementation of ML applications. One of the main limitations includes the varying quality of each extracted study in terms of how the MLOps implementation was presented. Hence, it was difficult to verify the presence and discuss in depth all steps of the MLOps workflow for each study. Furthermore, due to the inherent nature of a scoping review protocol, there may be a compromise on an in-depth discussion of each step within the MLOps workflow.

IoT data analytics in dynamic environments: From an automated machine learning perspective

Shear viscosity, though being a fundamental property of all fluids, is computationally expensive to calculate from equilibrium molecular dynamics simulations. Recently, machine learning (ML) methods have been used to augment molecular simulations in many contexts, thus showing promise to estimate viscosity too in a relatively inexpensive manner. However, ML methods face significant challenges—such as overfitting, when the size of the data set is small, as is the case with viscosity. In this work, we train seven ML models to predict the shear viscosity of a Lennard–Jones fluid, with particular emphasis on addressing issues arising from a small data set. Specifically, the issues related to model selection, performance estimation and uncertainty quantification were investigated. First, we show that the widely used performance estimation procedure of using a single unseen data set shows a wide variability—in estimating the errors on—small data sets. In this context, the common practice of using cross validation (CV) to select the hyperparameters (model selection) can be adapted to estimate the generalization error (performance estimation) as well. We compare two simple CV procedures for their ability to do both model selection and performance estimation, and find that k-fold CV based procedure shows a lower variance of error estimates. Also, these CV procedures naturally lead to an ensemble of trained ML models. We discuss the role of performance metrics in training and evaluation and propose a method to rank the ML models based on multiple metrics. Finally, two methods for uncertainty quantification—Gaussian process regression (GPR) and ensemble method—were used to estimate the uncertainty on individual predictions. The uncertainty estimates from GPR were also used to construct an applicability domain using which the ML models provided even more reliable predictions on an independent viscosity data set generated in this work. Overall, the procedures prescribed in this work, together, lead to robust ML models for small data sets.

BackgroundAccurate and interpretable models are essential for clinical decision-making, where predictions can directly impact patient care. Machine learning (ML) survival methods can handle complex multidimensional data and achieve high accuracy but require post-hoc explanations. Traditional models such as the Cox Proportional Hazards Model (Cox-PH) are less flexible, but fast, stable, and intrinsically transparent. Moreover, ML does not always outperform Cox-PH in clinical settings, warranting a diligent model validation. We aimed to develop a set of R functions to help explore the limits of Cox-PH compared to the tree-based and deep learning survival models for clinical prediction modelling, employing ensemble learning and nested cross-validation. MethodsWe developed a set of R functions, publicly available as the package “survcompare”. It supports Cox-PH and Cox-Lasso, and Survival Random Forest (SRF) and DeepHit are the ML alternatives, along with the ensemble methods integrating Cox-PH with SRF or DeepHit designed to isolate the marginal value of ML. The package performs a repeated nested cross-validation and tests for statistical significance of the ML’s superiority using the survival-specific performance metrics, the concordance index, time-dependent AUC-ROC and calibration slope.To get practical insights, we applied this methodology to clinical and simulated datasets with varying complexities and sizes. ResultsIn simulated data with non-linearities or interactions, ML models outperformed Cox-PH at sample sizes ≥ 500. ML superiority was also observed in imaging and high-dimensional clinical data. However, for tabular clinical data, the performance gains of ML were minimal; in some cases, regularised Cox-Lasso recovered much of the ML’s performance advantage with significantly faster computations. Ensemble methods combining Cox-PH and ML predictions were instrumental in quantifying Cox-PH’s limits and improving ML calibration. Traditional models like Cox-PH or Cox-Lasso should not be overlooked while developing clinical predictive models from tabular data or data of limited size. ConclusionOur package offers researchers a framework and practical tool for evaluating the accuracy-interpretability trade-off, helping make informed decisions about model selection.

Empirical software measurement is a dynamic research domain within software engineering. Measuring the software attributes during software development is crucial to ensure the successful delivery of a desired software product. It involves the measurement of multiple dynamic metrics based on software process and software product. The traditional approaches to measure the current software practices are not so appropriate due to the changing development process and increasing complexity. The changing environment requires techniques to make measurements that are adaptable to the changes. Machine learning (ML)-based techniques are prominently being deployed for measuring the software empirically. Past three decades witness the successful application of ML-based techniques for empirical measurements of software. ML is powerful enough to accurately measure the software attributes. But the synergy of ML algorithms and metaheuristic algorithm is apparently more promising in the domain of software measurement. Since 2001, metaheuristics are being fused with ML-based techniques to optimize the accuracy of the predictor based on ML. The estimation of software cost (or effort), estimation of development time (or schedule), prediction of software quality, and early detection of software anomalies are key tasks in the most popular research sector of entire field of software measurement. The vital role of metaheuristics along with ML in empirical software measurements is as follows. First, metaheuristics are used to optimize the hyperparameters of ML algorithms; second, to mitigate the problem of the curse of dimensionality of data; and third, to select the best ML model for the candidate problem. This study proposes a full-fledged description to the application of metaheuristic algorithms to empirical software measurements. The chapter begins with the introduction to the metaheuristic approach and its application to measure software empirically in conjunction with ML algorithms. Then, the state of the art of deployment of metaheuristics in the domain of software measurements is elaborated. Later, the study discusses the three major application areas of metaheuristics: hyperparameter tuning, dimensionality reduction, and model selection in detail. For each of these application areas, case studies are quoted for clearer insight into the techniques and their applications. The author proposes a novel metaheuristic technique in this chapter for feature selection. It brings more clarity for devising new methods. Then the chapter is concluded with special remarks on future scope and references made.

This article explores how machine learning techniques can be used to drive digital authentication to prevent fraud in digital payment technologies. First, it introduces the development trend and fraud risk of digital payment technology, and then analyzes the limitations of traditional authentication methods, focusing on the potential of machine learning in digital authentication. It then explores specific application scenarios of machine learning in digital authentication, including data collection and preparation, feature engineering, model selection and training, as well as real-time monitoring and anti-fraud processing. Finally, current challenges and solutions are discussed, as well as the future of machine learning in digital payment technology. Through in-depth analysis of these contents, the article aims to provide readers with valuable insights to help them better use machine learning technology to improve the security and reliability of digital payments and promote the sustainable development of the digital economy.

Up to 70% and even more of corporate Analytics Efforts fail!!! Even after these corporations have made very large investments, in time, talent, and money, in developing what they thought were good data and analytics programs. Why? Because the executives and decision makers and the entire analytics team have not considered the most important aspect of making these analytics efforts successful. In this Book II of "It’s All Analytics!" series, we describe two primary things: 1) What this "most important aspect" consists of, and 2) How to get this "most important aspect" at the center of the analytics effort and thus make your analytics program successful. This Book II in the series is divided into three main parts: Part I, Organizational Design for Success, discusses ……. The need for a complete company / organizational Alignment of the entire company and its analytics team for making its analytics successful. This means attention to the culture – the company culture culture!!! To be successful, the CEO’s and Decision Makers of a company / organization must be fully cognizant of the cultural focus on ‘establishing a center of excellence in analytics’. Simply, "culture – company culture" is the most important aspect of a successful analytics program. The focus must be on innovation, as this is needed by the analytics team to develop successful algorithms that will lead to greater company efficiency and increased profits. Part II, Data Design for Success, discusses ….. Data is the cornerstone of success with analytics. You can have the best analytics algorithms and models available, but if you do not have good data, efforts will at best be mediocre if not a complete failure. This Part II also goes further into data with descriptions of things like Volatile Data Memory Storage and Non-Volatile Data Memory Storage, in addition to things like data structures and data formats, plus considering things like Cluster Computing, Data Swamps, Muddy Data, Data Marts, Enterprise Data Warehouse, Data Reservoirs, and Analytic Sandboxes, and additionally Data Virtualization, Curated Data, Purchased Data, Nascent & Future Data, Supplemental Data, Meaningful Data, GIS (Geographic Information Systems) & Geo Analytics Data, Graph Databases, and Time Series Databases. Part II also considers Data Governance including Data Integrity, Data Security, Data Consistency, Data Confidence, Data Leakage, Data Distribution, and Data Literacy. Part III, Analytics Technology Design for Success, discusses …. Analytics Maturity and aspects of this maturity, like Exploratory Data Analysis, Data Preparation, Feature Engineering, Building Models, Model Evaluation, Model Selection, and Model Deployment. Part III also goes into the nuts and bolts of modern predictive analytics, discussing such terms as AI = Artificial Intelligence, Machine Learning, Deep Learning, and the more traditional aspects of analytics that feed into modern analytics like Statistics, Forecasting, Optimization, and Simulation. Part III also goes into how to Communicate and Act upon Analytics, which includes building a successful Analytics Culture within your company / organization. All-in-all, if your company or organization needs to be successful using analytics, this book will give you the basics of what you need to know to make it happen.

Anomaly detection is an important security mechanism for the Internet of Things (IoT). Existing works have been focused on developing accurate anomaly detection models. However, due to the resource-constrained nature of IoT networks and the requirement of real-time security operation, cost efficient (regarding computational efficiency and memory-consumption efficiency) approaches for anomaly detection are highly desirable in IoT applications. In this paper, we investigated machine learning (ML) enabled anomaly detection models for the IoT with regard to multi-objective optimization (Pareto optimization) that minimizes the detection error, execution time, and memory consumption simultaneously. Making use of well-known datasets consisting of network traffic traces captured in an IoT environment, we studied a variety of machine learning algorithms through the world-class H2O AI platform. Our experimental results show that the Gradient Boosting Machine, Random Forest, and Deep Learning models are the most accurate and fastest anomaly detection models; the Gradient Boosting Machine and Random Forest are the most accurate and memory-efficient models. These ML models form the Pareto-optimal set of anomaly detection models. Our results can be used by the industry to facilitate their selection of ML models for anomaly detection on various IoT networks based on their security requirements and system constraints.

Machine Learning Models Have Better Performance than Traditional Logistic Regression Models in Predicting the Risk of Diabetes

The increasing complexity of social science data and phenomena necessitates using advanced analytical techniques to capture nonlinear relationships that traditional linear models often overlook. This chapter explores the application of machine learning (ML) models in social science research, focusing on their ability to manage nonlinear interactions in multidimensional datasets. Nonlinear relationships are central to understanding social behaviors, socioeconomic factors, and psychological processes. Machine learning models, including decision trees, neural networks, random forests, and support vector machines, provide a flexible framework for capturing these intricate patterns. The chapter begins by examining the limitations of linear models and introduces essential machine learning techniques suited for nonlinear modeling. A discussion follows on how these models automatically detect interactions and threshold effects, offering superior predictive power and robustness against noise compared to traditional methods. The chapter also covers the practical challenges of model evaluation, validation, and handling imbalanced data, emphasizing cross-validation and performance metrics tailored to the nuances of social science datasets. Practical recommendations are offered to researchers, highlighting the balance between predictive accuracy and model interpretability, ethical considerations, and best practices for communicating results to diverse stakeholders. This chapter demonstrates that while machine learning models provide robust solutions for modeling nonlinear relationships, their successful application in social sciences requires careful attention to data quality, model selection, validation, and ethical considerations. Machine learning holds transformative potential for understanding complex social phenomena and informing data-driven psychology, sociology, and political science policy-making.

In-situ monitoring of Wire Arc Additive Manufacturing (WAAM) is crucial to provide real-time insights into the process for corrective feedback. Machine Learning (ML) approaches have proven to provide reasonable correlations between the input data and the process outputs in WAAM. However, there is a lack of knowledge with the development of an integrated ML development framework to address the simultaneous consideration of data processing and ML model development. This study aims to provide a search method to select ML frameworks with the combination of data processing and ML models with holistic performance metrics. Acoustic, welding current, welding fume, and the plate temperature datasets were collected during each printing operation. Subsequently, the collected datasets were preprocessed and used for feature extraction of the input data based on the intrinsic physical behavior of each time series dataset. The processed and extracted features were used to train six ML regression models to predict the corresponding measured height of the WAAM-printed workpieces. Mean Square Error (MSE), Mean Absolute Percentage Error (MAPE), and R-Squared metrics were then used to compare the performance of each of the regression models. Therefore, this study efficiently searches among different integrated ML frameworks to result in the combination of the data processing method and ML model with the best performance for the prediction of height in WAAM.

Software 2.0 refers to the fundamental shift in software engineering where using machine learning becomes the new norm in software with the availability of big data and computing infrastructure. As a result, many software engineering practices need to be rethought from scratch where data becomes a first-class citizen, on par with code. It is well known that 80--90% of the time for machine learning development is spent on data preparation. Also, even the best machine learning algorithms cannot perform well without good data or at least handling biased and dirty data during model training. In this tutorial, we focus on data collection and quality challenges that frequently occur in deep learning applications. Compared to traditional machine learning, there is less need for feature engineering, but more need for significant amounts of data. We thus go through state-of-the-art data collection techniques for machine learning. Then, we cover data validation and cleaning techniques for improving data quality. Even if the data is still problematic, hope is not lost, and we cover fair and robust training techniques for handling data bias and errors. We believe that the data management community is well poised to lead the research in these directions. The presenters have extensive experience in developing machine learning platforms and publishing papers in top-tier database, data mining, and machine learning venues.

Fraud detection in healthcare claims using machine learning: A systematic review.

With the urbanization process, an increasing number of sensors are being deployed in transportation systems, leading to an explosion of big data. To harness the power of this vast transportation data, various machine learning (ML) and artificial intelligence (AI) methods have been introduced to address numerous transportation challenges. However, these methods often require significant investment in data collection, processing, storage, and the employment of professionals with expertise in transportation and ML. Additionally, privacy issues are a major concern when processing data for real-world traffic control and management. To address these challenges, the research team proposes an innovative multi-agent framework called Independent Mobility Generative Pre-Trained Transformers (IDM-GPT) based on large language models (LLMs) for customized traffic analysis, management suggestions, and privacy preservation. IDM-GPT efficiently connects users, transportation databases, and ML models economically. IDM-GPT trains, customizes, and applies various LLM-based AI agents for multiple functions, including user query comprehension, prompts optimization, data analysis, model selection, and performance evaluation and enhancement. With IDM-GPT, users without any background in transportation or ML can efficiently and intuitively obtain data analysis and customized suggestions in near real-time based on their questions. Experimental results demonstrate that IDM-GPT delivers satisfactory performance across multiple traffic-related tasks, providing comprehensive and actionable insights that support effective traffic management and urban mobility improvement.