一、The learning problem
1. Course Introduction
1.2 What is Machine Learning
1.3 Applications of Machine Learning
1.4 Components of Machine Learning
1.5 Machine Learning and Other Fields
1.5.1 ML VS DM (Data Mining)
1.5.2 M L VS AI (artificial intelligence)
1.5.3 ML VS statistic
二、Learning to Answer Yes/No
2.1 Perceptron Hypothesis Set
2.2 Perceptron Learning Algorithm (PLA)
2.3 Guarantee of PLA
2.4 Non-Separable Data
三、Types of Learning
3.1 Learning with Different Output Space
3.1.1 binary classification
3.1.2 Multiclass Classification
3.1.3 Regression
3.1.4 Structured Learning
3.2 Learning with Different Data Label
3.2.1 Supervised Learning
3.2.2 Unsupervised Learning
3.2.3 Semi-supervised Learning
3.2.4 Reinforcement Learning
3.3 Learning with Different Protocol
3.4 Learning with Different Input Space
四、Feasibility of Learning
4.1 Learning is Impossible
4.2 Probability to the Rescue
4.3 Connection to Learning
4.4 Connection to Real Learning
五、Training versus Testing
5.1 Recap and Preview
5.2 Effective Number of Lines
5.3 Effective Number of Hypotheses
5.4 Break Point
六、Theory of Generalization
6.1 Restriction of Break Point
6.2 Bounding Function- Basic Cases
6.3 Bounding Function- Inductive Cases
6.4 A Pictorial Proof
七、The VC Dimension
7.1 Definition of VC Dimension
7.2 VC Dimension of Perceptrons
7.3 Physical Intuition of VC Dimension
7.4 Interpreting VC Dimension
八、Noise and Error
8.1 Noise and Probabilistic Target
8.2 Error Measure
8.3 Algorithmic Error Measure
8.4 Weighted Classification
九、Linear Regression
9.1 Linear Regression Problem
9.2 Linear Regression Algorithm
9.3 Generalization Issue
9.4 Linear Regression for Binary Classification
十、Logistic Regression
10.1 Logistic Regression Problem
10.2 Logistic Regression Error
10.3 Gradient of Logistic Regression Error
10.4 Gradient Descent
十一、Linear Models for Classification
11.1 Linear Models for Binary Classification
11.2 Stochastic Gradient Descent
11.3 Multiclass via Logistic Regression
11.4 Multiclass via Binary Classification
十二、Nonlinear Transformation
12.1 Quadratic Hypotheses
12.2 Nonlinear Transform
12.3 Price of Nonlinear Transform
12.4 Structured Hypothesis Sets
十三、Hazard of Overfitting
13.1 What is Overfitting?
13.2 The Role of Noise and Data Size
13.3 Deterministic Noise
13.4 Dealing with Overfitting
十四、Regularization
14.1 Regularized Hypothesis Set
14.2 Weight Decay Regularization
14.3 Regularization and VC Theory
14.4 General Regularizers
Word 书签
OLE_LINK3
OLE_LINK9