This repository holds the lecture notes for the class. Be aware that I may update the lecture notes after they are posted. Periodically check the most recent commit date and comment to see if you have the most recent version. A better strategy would be the use github to update your local version of the repository (e.g. git clone the repository and then update it with git pull).
I will try to keep this updated as to the notes/slides we cover for each day:
| Date | Slides |
|---|---|
| Aug. 26 | We went over the syllabus and gave a high level overview of what the course will entail. Additionally, we defined some common terms/concepts that will be used throughout the course. |
| Aug. 28 | We discussed missing data and imputation. We talked about some unsupervised imputation schemes and start to explore R Markdown via a simulation involving missing data and imputation |
| Sep. 02 | We covered aspects of the multivariate normal distribution. |
| Sep. 04 | We began talking about graphical models. We discussed covariance and correlation graphs as well as inference using the Wishart distribution. |
| Sep. 09 | We discussed partial correlation graphical models and inference using the bootstrap. We then introduced PCA. |
| Sep. 11 | We talked about two related methods of generalizing PCA: MDS and KPCA. MDS operates on the pairwise dissimilarities which KPCA nonlinearizes PCA by inserting a general kernel in for the Euclidean inner product. |
| Sep. 16 | We began looking at clustering by covering K-means clustering and beginning to talk about Hierarchical clustering |
| Sep. 18 | Finished talking about Hierarchical clustering and briefly discussed density-based (i.e. soft) clustering |
| Sep. 23 | We began discussing supervised learning, defining risk, loss, and Bayes' rules |
| Sep. 25 | We discuss various estimators of the Risk |
| Sep. 30 | Stepwise selection procedures: global and local solutions to a non-convex optimization problem |
| Oct. 02 | Ridge regression |
| Oct. 07 | Lasso regression |
| Oct. 09 | In class exercise on coordinate descent for regression and the lasso. Introduce elastic net and refitted lasso |
| Oct. 14 | Midterm Review |
| Oct. 16 | Relaxed lasso and in class exercise on batch gradient descent [Midterm from 6:00 pm to 8:00 pm in BLOC 149] |
| Oct. 21 | Defining risk for classification and logistic regression |
| Oct. 23 | Linear classifiers and the logistic elastic net |
| Oct. 28 | Evaluating classifiers |
| Oct. 30 | Nonparametric methods |
| Nov. 04 | Decision Trees |
| Nov. 06 | Bagging and random forest |
| Nov. 11 | |
| Nov. 13 | |
| Nov. 18 | |
| Nov. 20 | |
| Nov. 25 | |
| Nov. 27 | Thanksgiving, no class |
| Dec. 02 | Friday schedule, no class |
| Dec. 04 | Review for Final |
| Dec. 09 | Final, 3:30 to 5:30 pm |
Note The final is determined by the university and cannot be changed (https://registrar.tamu.edu/Courses,-Registration,-Scheduling/Final-Examination-Schedules#5-December11(Wednesday))
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent