About the Book
Machine Learning: A Bayesian and Optimization Perspective, 2nd edition, gives a unified perspective on machine learning by covering both pillars of supervised learning, namely regression and classification. The book starts with the basics, including mean square, least squares and maximum likelihood methods, ridge regression, Bayesian decision theory classification, logistic regression, and decision trees. It then progresses to more recent techniques, covering sparse modelling methods, learning in reproducing kernel Hilbert spaces and support vector machines, Bayesian inference with a focus on the EM algorithm and its approximate inference variational versions, Monte Carlo methods, probabilistic graphical models focusing on Bayesian networks, hidden Markov models and particle filtering. Dimensionality reduction and latent variables modelling are also considered in depth.
This palette of techniques concludes with an extended chapter on neural networks and deep learning architectures. The book also covers the fundamentals of statistical parameter estimation, Wiener and Kalman filtering, convexity and convex optimization, including a chapter on stochastic approximation and the gradient descent family of algorithms, presenting related online learning techniques as well as concepts and algorithmic versions for distributed optimization.
Focusing on the physical reasoning behind the mathematics, without sacrificing rigor, all the various methods and techniques are explained in depth, supported by examples and problems, giving an invaluable resource to the student and researcher for understanding and applying machine learning concepts. Most of the chapters include typical case studies and computer exercises, both in MATLAB and Python.
The chapters are written to be as self-contained as possible, making the text suitable for different courses: pattern recognition, statistical/adaptive signal processing, statistical/Bayesian learning, as well as courses on sparse modeling, deep learning, and probabilistic graphical models.
New to this edition:
Complete re-write of the chapter on Neural Networks and Deep Learning to reflect the latest advances since the 1st edition. The chapter, starting from the basic perceptron and feed-forward neural networks concepts, now presents an in depth treatment of deep networks, including recent optimization algorithms, batch normalization, regularization techniques such as the dropout method, convolutional neural networks, recurrent neural networks, attention mechanisms, adversarial examples and training, capsule networks and generative architectures, such as restricted Boltzman machines (RBMs), variational autoencoders and generative adversarial networks (GANs).
Expanded treatment of Bayesian learning to include nonparametric Bayesian methods, with a focus on the Chinese restaurant and the Indian buffet processes.
Table of Contents:
1. Introduction
2. Probability and stochastic Processes
3. Learning in parametric Modeling: Basic Concepts and Directions
4. Mean-Square Error Linear Estimation
5. Stochastic Gradient Descent: the LMS Algorithm and its Family
6. The Least-Squares Family
7. Classification: A Tour of the Classics
8. Parameter Learning: A Convex Analytic Path
9. Sparsity-Aware Learning: Concepts and Theoretical Foundations
10. Sparsity-Aware Learning: Algorithms and Applications
11. Learning in Reproducing Kernel Hilbert Spaces
12. Bayesian Learning: Inference and the EM Algorithm
13. Bayesian Learning: Approximate Inference and nonparametric Models
14. Montel Carlo Methods
15. Probabilistic Graphical Models: Part 1
16. Probabilistic Graphical Models: Part 2
17. Particle Filtering
18. Neural Networks and Deep Learning
19. Dimensionality Reduction and Latent Variables Modeling
About the Author :
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.
Review :
"This is a very complete Machine Learning book, as it covers statistical learning theory, both from frequentist and Bayesian perspectives. It also encompasses signal processing, probabilistic graphical models, deep learning, and latent variable modeling. It balances mathematical rigor with insightful comments to ease clear interpretation. The many examples make the text even more comprehensive. Each chapter has a well-curated list of references for further deepening on specific topics. Thus, it provides a thorough background for Machine Learning at an upper undergraduate level course. This book is also an excellent reference for practitioners to understand the necessary theory to apply Machine Learning with informed criteria." --Hamed Yazdanpanah, Postdoctoral Researcher, University of São Paulo
Reviews of the previous edition:
"Overall, this text is well organized and full of details suitable for advanced graduate and postgraduate courses, as well as scholars..." --Computing Reviews
"Machine Learning: A Bayesian and Optimization Perspective, Academic Press, 2105, by Sergios Theodoridis is a wonderful book, up to date and rich in detail. It covers a broad selection of topics ranging from classical regression and classification techniques to more recent ones including sparse modeling, convex optimization, Bayesian learning, graphical models and neural networks, giving it a very modern feel and making it highly relevant in the deep learning era. While other widely used machine learning textbooks tend to sacrifice clarity for elegance, Professor Theodoridis provides you with enough detail and insights to understand the "fine print". This makes the book indispensable for the active machine learner." --Prof. Lars Kai Hansen, DTU Compute - Dept. Applied Mathematics and Computer Science Technical University of Denmark
"Before the publication of Machine Learning: A Bayesian and Optimization Perspective, I had the opportunity to review one of the chapters in the book (on Monte Carlo methods). I have published actively in this area, and so I was curious how S. Theodoridis would write about it. I was utterly impressed. The chapter presented the material with an optimal mix of theoretical and practical contents in very clear manner and with information for a wide range of readers, from newcomers to more advanced readers. This raised my curiosity to read the rest of the book once it was published. I did it and my original impressions were further reinforced. S. Theodoridis has a great capability to disentangle the important from the unimportant and to make the most of the used space for writing. His text is rich with insights about the addressed topics that are not only helpful for novices but also for seasoned researchers. It goes without saying that my department adopted his book as a textbook in the course on machine learning." --Petar M. Djuric, Ph.D. SUNY Distinguished Professor Department of Electrical and Computer Engineering Stony Brook University, Stony Brook, USA
"As someone who has taught graduate courses in pattern recognition for over 35 years, I have always looked for a rigorous book that is current and appealing to students with widely varying backgrounds. The book on Machine Learning by Sergios Theodoridis has struck the perfect balance in explaining the key (traditional and new) concepts in machine learning in a way that can be appreciated by undergraduate and graduate students as well as practicing engineers and scientists. The chapters have been written in a self-consistent way, which will help instructors to assemble different sections of the book to suit the background of students" --Rama Cellappa, Distinguished University Professor, Minta Martin Professor of Engineering, Chair, Department of Electrical and Computer Engineering, University of Maryland, USA
