
- Home
- Books
- Artificial Intelligence & Machine Learning
- Machine Learning Foundations
- Foundations of Machine Learning, second edition
Foundations of Machine Learning by Mehryar Mohri is a graduate-level textbook on statistical learning theory and computational learning guarantees. Understanding algorithm behavior requires rigorous analysis of generalization bounds, sample complexity, and convergence properties. Mohri and his co-authors present core mathematical tools that connect empirical risk minimization with theoretical guarantees across supervised and online learning.
What you will learn
- Derive generalization error bounds using Rademacher complexity, growth functions, and VC-dimension
- Analyze the Probably Approximately Correct (PAC) framework and sample complexity of hypothesis classes
- Formulate support vector machines, margin bounds, and dual optimization problems via convex analysis
- Construct reproducing kernel Hilbert spaces (RKHS) and evaluate kernel substitution techniques
- Prove convergence and generalization bounds for AdaBoost and ensemble boosting algorithms
- Implement online learning algorithms, regret minimization strategies, and multi-armed bandit models
- Evaluate theoretical guarantees for ranking algorithms, multi-class classification, and regression problems
Who this is for
Graduate students, research scientists, and quantitative algorithmic engineers who seek mathematical proofs and theoretical foundations of machine learning. A solid background in linear algebra, real analysis, and probability theory is recommended.


