As machine learning applications gain widespread adoption and integration in a variety of applications, including safety and mission-critical systems, the need for robust evaluation methods grows more urgent. This book compiles scattered information on the topic from research papers and blogs to provide a centralized resource that is accessible to students, practitioners, and researchers across the sciences. The book examines meaningful metrics for diverse types of learning paradigms and applications, unbiased estimation methods, rigorous statistical analysis, fair training sets, and meaningful explainability, all of which are essential to building robust and reliable machine learning products. In addition to standard classification, the book discusses unsupervised learning, regression, image segmentation, and anomaly detection. The book also covers topics such as industry-strength evaluation, fairness, and responsible AI. Implementations using Python and scikit-learn are available on the book’s website.
Editorial Reviews
Review
‘By its nature, machine learning has always had evaluation at its heart. As the authors of this timely and important book note, the importance of doing evaluation properly is only increasing as we enter the age of machine learning deployment. The book showcases Japkowicz’ and Boukouvalas’ encyclopaedic knowledge of the subject as well as their accessible and lucid writing style. Quite simply required reading for machine learning researchers and professionals.’ Peter Flach, University of Bristol
‘I recommend this book for students and instructors of machine learning, both traditional and deep. The authors state: ‘The purpose of this book is to present a concise, yet complete, intuitive, yet formal, presentation of machine learning evaluation.’ In this important and useful book, they succeed on all counts.’ Creed Jones, Computing Reviews
Book Description
A practical guide to robust performance evaluation methods machine learning models for modern industrial-strength applications.
About the Author
Nathalie Japkowicz is Professor and Chair of the Department of Computer Science at American University, Washington DC. She previously taught at the University of Ottawa. Her current research focuses on lifelong anomaly detection and hate speech detection. In the past, she researched one-class learning and the class imbalance problem extensively. She has received numerous awards, including Test of Time and Distinguished Service awards.
Zois Boukouvalas is Assistant Professor in the Department of Mathematics and Statistics at American University, Washington DC. His research focuses on the development of interpretable multi-modal machine learning algorithms, and he has been the lead principal investigator of several research grants. Through his research and teaching activities, he is creating environments that encourage and support the success of underrepresented students for entry into machine learning careers.
Wow! eBook
