| Date | Content | Reading | Slides |
|---|---|---|---|
| Intro, Maximum Likelihood, Linear Regression, Overfitting, Regularization | |||
| 1/3 M | Welcome/overview Maximum likelihood for Bernoulli |
Murphy 1.1 - 1.3, 2.2-2.6, 4.1, 6.4 Optional: HTF 1, Wasserman 9.1.-9.8 | Combined week 1 slides, Lecture 1 slides, ( pre-annotated , live-annotated ) |
| 1/5 W | Maximum likelihood for Gaussian, Linear least squares | Murphy 6.4, 7.1-7.3
Optional: HTF 3.1-3.2 |
Lecture 2 slides, ( pre-annotated , live-annotated ) |
| 1/7 F | Linear least squares | Murphy 6.4, 7.1-7.3
Optional: HTF 3.1-3.2 |
Lecture 3 slides, ( pre-annotated , live-annotated ) |
| 1/10 M | Generalized linear regression, Bias-Variance tradeoff |
Murphy 6.4, 7.5.1, Fortmann-Roe essay Optional: HTF 7.1-7.4 |
Combined week 2 slides, Lecture 4 slides, ( pre-annotated , live-annotated ), demo2_lin.ipynb demo3_diabetes.ipynb |
| 1/12 W | Bias-Variance tradeoff |
Murphy 6.4, 7.5.1 Optional: HTF 7.1-7.4 |
Lecture 5 slides, ( pre-annotated , live-annotated ) |
| 1/14 F | Bias-Variance tradeoff and Overfitting |
Murphy 6.4-6.5 Optional: HTF 7.1-7.4 |
Lecture 6 slides, ( pre-annotated , live-annotated ) demo4_tradeoff.html, demo4_tradeoff.ipynb, lecture2_polynomialfit.ipynb |
| 1/17 M | Martin Luther King day | ||
| 1/19 W | Ridge regression |
Murphy 6.4-6.5 Optional: HTF 7.10-7.12, 3.4 |
Combined week 3 slides, Lecture 7 slides, ( live-annotated ) demo5_ridge.ipynb , lecture3_ridge.ipynb , |
| 1/21 F | k-fold cross validation |
Murphy 6.4-6.5, Optional: HTF 3 |
Lecture 8 slides, ( pre-annotated , live-annotated ) |
| 1/24 M | Lasso |
Murphy 13.1, 13.3 - 13.4.1 Optional: HTF 3 |
Combined week 4 slides, Lecture 9 slides ( live-annotated ) demo_lasso.html , demo_lasso.ipynb , |
| Classification, Optimization | |||
| 1/26 W | Convexity |
Murphy 8.1-8.3, 8.5 Roughgarden-Valiant notes Optional: Gradient descent algorithms by Ruder Optional: Zen of gradient descent by Moritz Hardt Optional: HTF 4.1-4.2, 4.4 |
Lecture 10 slides (live-annotated ) |
| 1/28 F | Classification, Logistic Regression |
Murphy 8.1-8.3, 8.5
Optional: HTF 4.1-4.2, 4.4 |
Lecture 11 slides (pre-annotated ,live-annotated ), demo8_classification.ipynb |
| 1/31 M | Gradient Descent |
Daume 4, 11, Murphy 8.5.0 Murphy 14.5 Optional: HTF 4.5, 12-12.2; Optional: HTF 4.1-4.3.1, 18.7; EH 2-2.2, 10-10.4, 11-11.2 |
Combined week 5 slides, Lecture 12 slides (pre-annotated ,live-annotated), demo7_RobustGD.ipynb |
| 2/2 W | Stochastic gradient descent |
Daume 4, 11, Murphy 8.5.0, 8.5.4
Murphy 14.5
Optional: HTF 4.5, 12-12.2;
Optional: HTF 4.1-4.3.1, 18.7; EH 2-2.2, 10-10.4, 11-11.2 |
Lecture 13 slides (pre-annotated ,live-annotated) |
| 2/4 F | Coordinate descent |
Daume 4, 11, Murphy 8.5.0, 8.5.4
Murphy 14.5
Optional: HTF 4.5, 12-12.2;
Optional: HTF 4.1-4.3.1, 18.7; EH 2-2.2, 10-10.4, 11-11.2 |
Lecture 14 slides (pre-annotated,live-annotated) |
| 2/7 M | Support Vector Machine (SVM) |
Daume 4, 11, Murphy 8.5.0, 8.5.4
Murphy 14.5
Optional: HTF 4.5, 12-12.2;
Optional: HTF 4.1-4.3.1, 18.7; EH 2-2.2, 10-10.4, 11-11.2 |
Combined week 6 slides, Lecture 15 slides (pre-annotated,live-annotated) |
| Non-linear models | |||
| 2/9 W | Kernels | Lecture 16 slides (live-annotated) | |
| 2/11 F | kernels | Lecture 17 slides (live-annotated) | |
| 2/14 M | Bootstrap, Neural Network |
Murphy 28
Optional: HTF 11 Dive into Deep Learning, Zhang et al https://playground.tensorflow.org/ |
Combined week 7 slides, Lecture 18 slides (live-annotated) |
| 2/16 W | Neural Network |
Murphy 28
Optional: HTF 11 Dive into Deep Learning, Zhang et al https://playground.tensorflow.org/ |
Lecture 19 slides (live-annotated) |
| 2/18 F | Neural Network |
Murphy 28
Optional: HTF 11 Dive into Deep Learning, Zhang et al https://playground.tensorflow.org/ |
Lecture 20 slides (live-annotated) |
| 2/21 M | President's day | ||
| 2/23 W | Nearest Neighbor | HTF 2-2.5, HTF 6-6.3, EH 10-10.4, EH 11-11.2 | Combined week 8 slides, Lecture 21 slides (live-annotated) |
| Unsupervised Learning | |||
| 2/25 F | Principal Componenet Analysis |
HTF 2.3-2.5, 7.11, 8.2 Murphy 6.1-6.3, 12.1.0, 12.2 Daume, chapter 15 Optional: HTF 5.9, 12.3; 14.5 PCA note 1 and notes 2, Matrix completion paper analyzing gradient descent paper , Influential paper on deanonymizing netflix challenge using IMDB paper |
Lecture 22 slides (live-annotated) |
| 2/28 M | PCA and autoencoder | Combined week 9 slides, Lecture 23 slides (live-annotated) | |
| 3/2 W | k-means and Gaussian mixtures |
HTF 14.3.6-14.3.7 Murphy 11.2-11.4, k-means ++ paper |
Lecture 24 slides (live-annotated) |
| 3/4 F | Spectral clistering, Generative Adversarial Networks | Lecture 25 slides (live-annotated) | |
| 3/7 M | Fairness in AI (Guest lecture by Chase King) | Lecture 26 slides | |
| 3/9 W | Generative Adversarial Networks | Lecture 27 slides | |
| 3/11 F | Robust Machine Learning | Lecture 28 slides |