课时1：Course Introduction
课时2：What is Machine Learning
课时3：Applications of Machine Learning
课时4：Components of Machine Learning
课时5：Machine Learning and Other Fields
课时6：Perceptron Hypothesis Set
课时7：Perceptron Learning Algorithm (PLA)
课时8：Guarantee of PLA
课时9：Non-Separable Data
课时10：Learning with Different Output Space
课时11：Learning with Different Data Label
课时12：Learning with Different Protocol
课时13：Learning with Different Input Space
课时14：Learning is Impossible
课时15：Probability to the Rescue
课时16：Connection to Learning
课时17：Connection to Real Learning
课时18：Recap and Preview
课时19：Effective Number of Lines
课时20：Effective Number of Hypotheses
课时21：Break Point
课时22：Restriction of Break Point
课时23：Bounding Function- Basic Cases
课时24：Bounding Function- Inductive Cases
课时25：A Pictorial Proof
课时26：Definition of VC Dimension
课时27：VC Dimension of Perceptrons
课时28：Physical Intuition of VC Dimension
课时29：Interpreting VC Dimension
课时30：Noise and Probabilistic Target
课时31：Error Measure
课时32：Algorithmic Error Measure
课时33：Weighted Classification
课时34：Linear Regression Problem
课时35：Linear Regression Algorithm
课时36：Generalization Issue
课时37：Linear Regression for Binary Classification
课时38：Logistic Regression Problem
课时39：Logistic Regression Error
课时40：Gradient of Logistic Regression Error
课时41：Gradient Descent
课时42：Linear Models for Binary Classification
课时43：Stochastic Gradient Descent
课时44：Multiclass via Logistic Regression
课时45：Multiclass via Binary Classification
课时46：Quadratic Hypothesis
课时47：Nonlinear Transform
课时48：Price of Nonlinear Transform
课时49：Structured Hypothesis Sets
课时50：What is Overfitting
课时51：The Role of Noise and Data Size
课时52：Deterministic Noise
课时53：Dealing with Overfitting
课时54：Regularized Hypothesis Set
课时55：Weight Decay Regularization
课时56：Regularization and VC Theory
课时57：General Regularizers
课时58：Model Selection Problem
课时59：Validation
课时60：Leave-One-Out Cross Validation
课时61：V-Fold Cross Validation
课时62：Occam-'s Razor
课时63：Sampling Bias
课时64：Data Snooping
课时65：Power of Three
