Mathematical Analysis of Machine Learning Algorithms

 The  mathematical theory of machine learning not only explains the current  algorithms but can also motivate principled approaches for the future.  This self-contained textbook introduces students and researchers of AI  to the main mathematical techniques used to analyze machine learning  algorithms, with motivations and applications. Topics covered include  the analysis of supervised learning algorithms in the iid setting, the  analysis of neural networks (e.g. neural tangent kernel and mean-field  analysis), and the analysis of machine learning algorithms in the  sequential decision setting (e.g. online learning, bandit problems, and  reinforcement learning). Students will learn the basic mathematical  tools used in the theoretical analysis of these machine learning  problems and how to apply them to the analysis of various concrete  algorithms. This textbook is perfect for readers who have some  background knowledge of basic machine learning methods, but want to gain  sufficient technical knowledge to understand research papers in  theoretical machine learning.