ML Book 5 - Support Vector Machines and Kernel Methods for Dummies

Table of Contents

1. Introduction

1.1 What are Support Vector Machines and Kernel Methods?

1.2 Why are they useful for data analysis?

1.3 How do they work?

1.4 What are the main challenges and limitations?

2. Basic Concepts and Terminology

2.1 Linear and Nonlinear Classification and Regression

2.2 Hyperplanes and Margins

2.3 Support Vectors and Dual Formulation

2.4 Kernels and Kernel Functions

2.5 Kernel Trick and Feature Mapping

3. Support Vector Machines for Classification

3.1 Hard and Soft Margin SVM

3.2 Regularization and Cross-Validation

3.3 Multiclass Classification and One-vs-One and One-vs-All Strategies 3.4 SVM Algorithms and Implementations

4. Support Vector Machines for Regression

4.1 Epsilon-Insensitive Loss and Epsilon-Tube

4.2 Support Vector Regression and Cost Function

4.3 Nu-Support Vector Regression and Nu-Parameter

4.4 SVM Algorithms and Implementations

5. Kernel Methods for Data Analysis

5.1 Kernel Principal Component Analysis and Dimensionality Reduction 5.2 Kernel K-Means and Clustering

5.3 Kernel Ridge Regression and Regularized Least Squares

5.4 Kernel Algorithms and Implementations

6. Advanced Topics and Applications

6.1 Non-Standard Kernels and Kernel Design

6.2 Multiple Kernel Learning and Kernel Alignment

6.3 Support Vector Machines and Kernel Methods for Structured Data

6.4 Support Vector Machines and Kernel Methods for Text, Image, and Audio Data