Lectures & Readings
Most course readings are taken from Machine Learning: A Probabilistic Perspective (MLaPP), a draft textbook in preparation by Prof. Kevin Murphy. The first chapter is freely available online. Later chapters will be distributed via a reader available from the Metcalf Copy Center.
| Date | Topic | Primary Readings | Additional Readings | Materials |
|---|---|---|---|---|
| 1/27 | Course Overview | MLaPP: 1.1 | slides | |
| 2/01 | Classification Probability: What and Why? |
MLaPP: 1.2 MLaPP: 2.2 |
slides | |
| 2/03 | Discrete Random Variables Maximum Likelihood & Naive Bayes |
MLaPP: 2.3, 2.6 MLaPP: 1.4.3, 3.1, 3.3 |
slides | |
| 2/08 | Categorical Maximum Likelihood Continuous Random Variables Bayesian Inference |
MLaPP: 3.3-3.5 MLaPP: 2.4, 2.5 MLaPP: 4.1-4.2 |
slides | |
| 2/10 | Gaussian Maximum Likelihood Categorical Bayesian Estimation Decision Theory, ROC Curves |
MLaPP: 3.2 MLaPP: 1.8, 4.5, 4.6 MLaPP: 8.1-8.3 |
slides | |
| 2/15 | Decision Theory, Precision & Recall Model Selection, Cross-validation |
MLaPP: 8.2, 8.3 MLaPP: 1.8, 4.8, 4.9 |
slides | |
| 2/17 | Multivariate Gaussian Distributions Discriminant Analysis |
MLaPP: 5.1-5.3 MLaPP: 1.4.1 |
slides | |
| 2/22 | Brown Holiday: No Lecture | |||
| 2/24 | Linear Regression & Least Squares Bayesian Linear Regression |
MLaPP: 1.3, 3.7 MLaPP: 5.3-5.5 |
slides | |
| 3/01 | Bayesian Regression & Prediction Bayesian Model Selection Logistic Regression |
MLaPP: 5.5 MLaPP: 1.8.5, 4.8 MLaPP: 1.2.7-1.2.13 |
slides | |
| 3/03 | Logistic Regression Gradient Descent, Newton's Method |
MLaPP: 3.8, 14.2-14.3 MLaPP: 10.1-10.3 |
slides | |
| 3/08 | Iteratively Reweighted Least Squares Stochastic Gradient Descent |
MLaPP: 10.3 MLaPP: 10.2 |
slides | |
| 3/10 | Exponential Families Generalized Linear Models Robust Linear Regression |
MLaPP: 33.5 MLaPP: 14.5-14.6 MLaPP: 1.3.4, 13.3 |
slides | |
| 3/15 | Midterm Exam: In Class | |||
| 3/17 | Binary Feature Selection & Search L1 Regularization & Sparsity |
MLaPP: 15.2 MLaPP: 15.3 |
slides | |
| 3/22 | L1 Optimization & Sparsity Kernel Methods Gaussian Process Regression |
MLaPP: 15.4 MLaPP: 16.1-16.2 MLaPP: 16.3 |
slides | |
| 3/24 | Gaussian Process Classification, GLMs Laplace Approximations |
MLaPP: 16.3-16.4 MLaPP: 11.2 |
slides | |
| 3/29 | Spring Break: No Lecture | |||
| 3/31 | Spring Break: No Lecture | |||
| 4/05 | Perceptron Algorithm GP Laplace Approximations |
MLaPP: 10.2.4 MLaPP: 16.4 |
slides | |
| 4/07 | Margins & Support Vector Machines Clustering & K-Means Algorithm |
MLaPP: 17.4 MLaPP: 1.5.1, 19.1, 19.2 |
slides | |
| 4/12 | Probabilistic Mixture Models EM for Gaussian Mixtures |
MLaPP: 1.5.1, 19.3 MLaPP: 10.4.1-10.4.2 |
slides | |
| 4/14 | Expectation Maximization Algorithm | MLaPP: 10.4 | slides | |
| 4/19 | Principal Components Analysis Factor Analysis & Probabilistic PCA |
MLaPP: 1.5.2, 20.2-20.3 | slides | |
| 4/21 | EM for Factor Analysis & PCA | MLaPP: 20.2-20.3 | EM for PPCA | slides |
| 4/26 | Topic Models Hidden Markov Models |
MLaPP: 21.2, 21.4 MLaPP: 6.8, 22.2 |
slides topic models |
|
| 4/28 | IPP Symposium: No Lecture | |||
| 5/03 | Hidden Markov Models Monte Carlo, MCMC, Gibbs Samplers |
MLaPP: 22.2-22.3 MLaPP: 12.2, 12.5, 12.6 |
slides | |
| 5/05 | Gibbs Samplers Directed Graphical Models |
MLaPP: 12.6, 12.9 MLaPP: 6.2, 6.3, 6.6, 6.7 |
slides | |
| 5/10 | Final Exam Review Session | slides | ||
| 5/12 | Graduate Project Presentations |
Recitations
| Date | Topic | Presenters | Readings | Materials |
|---|---|---|---|---|
| 2/03 | Matlab Tutorial | Jason & Soumya | Matlab | |
| 2/10 | Continuous Random Variables Bayesian Estimation |
Jason | MLaPP: 4.5, 33 | Matlab notes |
| 2/17 | Linear Algebra Tutorial | Soumya | MLaPP: 32 | notes |
| 2/24 | Multivariate Gaussian Distributions | Jason | MLaPP: 5 | notes |
| 3/03 | Continuous Optimization | Soumya | MLaPP: 10.1-10.3, 31.4 | Matlab slides |
| 3/10 | Midterm Review Session | Jason & Soumya | MLaPP: 1-5, 8, 10 | slides |
| 3/17 | No Recitation | |||
| 3/24 | Lagrange Multipliers | Jason | MLaPP: 31.5, 31.8 | notes |
| 3/31 | No Recitation | |||
| 4/07 | Gaussian Processes | Soumya | MLaPP: 16 | notes slides |
| 4/14 | EM Algorithm | Jason | MLaPP: 10.4 | notes |
| 4/21 | Markov Chains | Soumya | MLaPP: 2.8, 31.7 | notes |
| 4/28 | Dynamic Programming & HMMs | Jason | MLaPP: 22.2 | notes slides |
| 5/05 | No Recitation |