Statistical Machine Learning Course

Lectures sourced from UC Berkeley STAT 154.

• Class introduction
• Linear regression in matrix notation (review)
• Linear regression in an ML setting
• Making new regressors from old
• Bias and variance tradeoff in prediction
• Ridge (L2) and Lasso (L1) Regression
• Generalization error and uniform laws
• Generalization and uniform laws
• VC dimension and uniform laws for zero–one loss
• Cross-validation
• Loss functions for classification
• Trading off false positives and negatives
• The Perceptron Algorithm
• Support Vector Machines
• Gradient descent and stochastic gradient descent
• Classification and Regression Trees (CART)
• Bagging and boosting
• Bagging and boosting
• The kernel trick
• Inner prodcuts and valid kernels
• Reproducing kernel Hilbert spaces
• Linear algebra review materials