These are the lecture notes from last year. Updated versions will be posted during the quarter.
Course information
Overview and examples
Predictors
Validation
Features
Empirical risk minimization
Constant predictors
Non-quadratic losses
House prices example and house.jl, houseplots.jl, house.csv
Non-quadratic regularizers
Neural networks
Classifiers
ERM for classifiers
Boolean classification
Multi-class classification
Probabilistic classification
ERM for probabilistic classification
Unsupervised learning
Principal components analysis
Optimization
Prox-gradient method
These notes will not be covered in the lecture videos, but you should read these in addition to the notes above.
Notation