Machine Learning

Sergios Theodoridis is professor emeritus of machine learning and data processing with the National and Kapodistrian University of Athens, Athens, Greece. He has also served as distinguished professor with the Aalborg University Denmark and as professor with the Chinese University of Hong Kong, Shenzhen, China. In 2023, he received an honorary doctorate degree (D.Sc) from the University of Edinburgh, U.K. He has also received a number of prestigious awards, including the 2014 IEEE Signal Processing Magazine Best Paper Award, the 2009 IEEE Computational Intelligence Society Transactions on Neural Networks Outstanding Paper Award, the 2017 European Association for Signal Processing (EURASIP) Athanasios Papoulis Award, the 2014 IEEE Signal Processing Society Carl Friedrich Gauss Education Award, and the 2014 EURASIP Meritorious Service Award. He has served as president of EURASIP and vice president for the IEEE Signal Processing Society. He is a Fellow of EURASIP and a Life Fellow of IEEE. He is the coauthor of the book Pattern Recognition, 4th edition, Academic Press, 2009 and of the book Introduction to Pattern Recognition: A MATLAB Approach, Academic Press, 2010.

• 1. Introduction2. Probability and Stochastic Processes3. Learning in Parametric Modelling: Basic Concepts and Directions4. Mean-Square Error Linear Estimation5. Stochastic Gradient Descent: the LMS Algorithm and its Family6. The Least-Squares Family7. Classification: A Tour of the Classics8. Parameter Learning: A Convex Analytic Path9. Sparsity-Aware Learning: Concepts and Theoretical Foundations10. Sparsity-Aware Learning: Algorithms and Applications11. Learning in Reproducing Kernel Hilbert Spaces12. Bayesian Learning: Inference and the EM Algorithm13. Bayesian Learning: Approximate Inference and Nonparametric Models14. Monte Carlo Methods15. Probabilistic Graphical Models: Part 116. Probabilistic Graphical Models: Part 217. Particle Filtering18. Neural Networks and Deep Learning: Part 119. Neural Networks and Deep Learning: Part 220. Dimensionality Reduction and Latent Variables Modeling

This excellent book offers a rare blend of breadth, depth, and clarity, serving equally well as a textbook, research monograph, and reference guide. It presents a unified view of modernMachine Learning by dedicating a full chapter to recent breakthroughs such as Transformers, Self-supervision, and Diffusion models, while also providing a solid foundation in classical topics including regression, classification, sparse modeling, kernel methods, Bayesian learning, and graphical models. Neural networks are introduced through their historical evolution, beginning with the perceptron and progressing through convolutional and recurrent architectures, generative adversarial networks, and variational autoencoders. The exposition balances conceptual insight with analytical precision and is enriched with case studies, examples, problems, and computational exercises that illuminate both theory and practice. Written by a distinguished author with deep expertise in the field, this book is a timely and indispensable resource for educators, students, and researchers who seek more than a black box treatment and want to understand the principles that drive advances in Machine Learning. - Georgios Giannakis, McKnight Presidential Chair, ECE Dept., University of Minnesota. Machine Learning (Third Edition) by Theodoridis provides a rigorous and conceptually unified treatment that situates modern ML methods within the broader framework of statistical inference, optimization, and probabilistic modeling. The text excels in its integration of classical pattern-recognition foundations with contemporary advances including variational inference, generative models, kernelized learning, and deep learning. It is a rare text that can serve simultaneously as a research companion, a teaching resource, and a bridge between statistical ML theory and practical algorithmic design. If you are a student looking for a machine-learning textbook that is clear, friendly, and genuinely helpful, Theodoridis’s Machine Learning (Third Edition) is an excellent choice. - Rama Chellapa, Bloomberg Distinguished Professor, Johns Hopkins UniversityThe book offers a comprehensive treatment of Machine Learning, ranging from the classics of classification and regression to modern deep learning. There is a detailed exposition of convex analysis, compressed sensing and sparsity-aware learning. Subsequently the book gets into Bayesian analysis and a detailed coverage of graphical models, providing an excellent exposition of classical unsupervised learning. The last part covers neural networks including modern research topics like GANs, Diffusions and Transformers. Overall, this is a comprehensive and insightful resource for anyone seeking depth and breadth in Machine Learning. - Alexandros G Dimakis, EECS, UC Berkeley