统计学习理论是针对小样本情况研究统计学习规律的理论，是传统统汁学的重要发展和补充，为研究有限样本情况下机器学习的理论和方法提供了理论框架，其核心思想是通过控制学习机器的容量实现对推广能力的控制。在这一理论中发展出的支持向量机方法是一种新的通用学习机器，较以往方法表现出很多理论和实践上的优势。本书是该领域的权威著作，由该领域的创立者来讲述统计学习理论的本质，着重介绍了统计学习理论和支持向量机的关键思想、结论和方法，以及该领域的新进展。

Preface to the Second Edition
Preface to the First Edition
Introduction: Four Periods in the Research of the Learning Problem
Rosenblatt's Perceptron (The 1960s)
Construction of the Fundamentals of Learning Theory(The 1960s–1970s)
Neural Networks (The 1980s)
Returning to the Origin (The 1990s)
Chapter 1 Setting of the Learning Problem
1.1 Function Estimation Model
1.2 The Problem of Risk Minimization
1.3 Three Main Learning Problems
1.3.1 Pattern Recognition
1.3.2 Regression Estimation
1.3.3 Density Estimation (Fisher–Wald Setting)
1.4 The General Setting of the Learning Problem
1.5 The Empirical Risk Minimization (ERM) Inductive Principle
1.6 The Four Parts of Learning Theory
1.7 The Classical Paradigm of Solving Learning Problems
1.7.1 Density Estimation Problem (MaximumLikelihood Method)
1.7.2 Pattern Recognition (Discriminant Analysis) Problem
1.7.3 Regression Estimation Model
1.7.4 Narrowness of the ML Method
1.8 Nonparametric Methods of Density Estimation
1.8.1 Parzen's Windows
1.8.2 The Problem of Density Estimation Is Ill-Posed
1.9 Main Principle for Solving Problems Using a Restricted Amount of Information
1.10 Model Minimization of the Risk Based on Empirical Data
1.10.1 Pattern Recognition
1.10.2 Regression Estimation
1.10.3 Density Estimation
1.11 Stochastic Approximation Inference
Chapter 2 Consistency of Learning Processes
2.1 The Classical Definition of Consistency and the Concept of Nontrivial Consistency
2.2 The Key Theorem of Learning Theory
2.2.1 Remark on the ML Method
2.3 Necessary and Sufficient Conditions for Uniform Two-Sided Convergence
2.3.1 Remark on Law of Large Numbers and Its Generalization
2.3.2 Entropy of the Set of Indicator Functions
2.3.3 Entropy of the Set of Real Functions
2.3.4 Conditions for Uniform Two-Sided Convergence
2.4 Necessary and Sufficient Conditions for Uniform One-Sided Convergence
2.5 Theory of Nonfalsifiability
2.5.1 Kant's Problem of Demarcation and Popper's Theory of Nonfalsifiability
2.6 Theorems on Nonfalsifiability
2.6.1 Case of Complete (Popper's) Nonfalsifiability
2.6.2 Theorem on Partial Nonfalsifiability
2.6.3 Theorem on Potential Nonfalsifiability
2.7 Three Milestones in Learning Theory Informal Reasoning and Comments
2.8 The Basic Problems of Probability Theory and Statistics
2.8.1 Axioms of Probability Theory
2.9 Two Modes of Estimating a Probability Measure
……
Chapter 3 Bounds on the Rate of Convergence ofLearning Processes
Chapter 4 Controlling the Generalization Ability of Learning Processes
Chapter 5 Methods of Pattern Recognition
Chapter 6 Methods of Function Estimation
Chapter 7 Direct Methods in Statistical Learning Theory
Chapter 8 The Vicinal Risk Minimization Principle and the SVMs
Chapter 9 Conclusion: What Is Important inLearning Theory?
References
Index