Schedule

Topics for each class day, as well as links to the pre-class videos and slides, are listed below.

Before each class, watch the pre-class video and answer the corresponding comprehension questions on Moodle.

Tentative overall schedule

Regression tasks: Weeks 2-5

  • Topics: model evaluation, model building, nonparametric methods, tools for modeling nonlinearity

Classification tasks: Weeks 6-9

  • Topics: model evaluation, logistic regression, tree-based methods, support vector methods

Unsupervised learning: Weeks 10-11

  • Topics: principal components analysis, clustering

Other topics + project work time: Weeks 12-14

  • Suggested topic: deep learning

Week 2: 1/28 - 2/1

  • Monday: Assumptions of linear regression (Video, Slides)
    Related ISLR reading: Section 2.1 gives some general background that is not particular to linear regression assumptions but sets up some key foundational ideas.
  • Friday: Model evaluation metrics for regression (Video, Slides)
    Related ISLR reading: Sections 2.2.1-2.2.2 talk about MSE, training and test data, test error, cross-validation, and the bias-variance tradeoff. The bias-variance tradeoff will come up a little later in class, but feel free to preview the ideas now.

Week 3: 2/4 - 2/8

  • Monday: Cross-validation (Video, Slides)
    Related ISLR reading: Section 5.1 is devoted to cross-validation. See also Sections 2.2.1-2.2.2.
  • Wednesday: We’ll finish reviewing cross validation, then talk briefly about variable selection methods for building models (Video, Slides)
    Related ISLR reading: Section 6.1
  • Friday: Quiz 1. Remaining time will be left for working on homework.

Week 4: 2/11 - 2/15

  • Monday: Shrinkage/regularization methods for model building (Video, Slides)
    Related ISLR reading: Section 6.2
  • Wednesday: Continue discussing shrinkage methods
  • Friday: K-nearest neighbors regression and the bias-variance tradeoff (Video, Slides)
    Related ISLR reading: Section 2.2.2 for the bias-variance tradeoff and Section 3.5 for K-nearest neighbors regression

Week 5: 2/18 - 2/22

  • Monday: Modeling nonlinear trends with natural splines (Video, Slides)
    Related ISLR reading: Sections 7.1-7.4
  • Wednesday: Local regression and generalized additive models (Video, Slides)
    Related ISLR reading: Sections 7.6-7.7
  • Friday: Quiz 2. Remaining time for a review activity.

Week 6: 2/25 - 3/1

  • Monday: Logistic regression (Video, Slides)
    Related ISLR reading: Sections 4.1-4.3
  • Wednesday: Finish up logistic regression. Using old tools in the classification setting
    Related ISLR reading: Section 7.7.2 (GAMs for Classification Problems), pages 39-42 (KNN for classification)
  • Friday: Decision trees (Video, Slides)
    Related ISLR reading: 8.1

Week 7: 3/4 - 3/8

  • Monday: Decision trees
  • Wednesday: Bagging and random forests (Video, Slides)
    Related ISLR reading: 8.2
  • Friday: Quiz 3. Finish up bagging and random forests.

Week 8: 3/11 - 3/15

Review & midterm exam

Happy Spring Break!

Week 9: 3/25 - 3/29

  • Monday: Review midterm exam, start thinking about final projects
  • Wednesday: Support vector machines (Video, Slides)
    Related ISLR reading: Sections 9.1-9.4
  • Friday: No class - MSCS Capstone Days

Week 10: 4/1 - 4/5

  • Monday: Continue with support vector machines
  • Wednesday: K-means clustering (Video, Slides)
    Related ISLR reading: Section 10.3 (specifically 10.3.1 for K-means)
  • Friday: Quiz 4. Hierarchical clustering (Video, Slides)
    Related ISLR reading: Section 10.3 (specifically 10.3.2 for hierarchical clustering)

Week 11: 4/8 - 4/12

  • Monday: Wrap up clustering
  • Wednesday: Principal components analysis (Video, Slides)
    Related ISLR reading: Section 10.2
  • Friday: Wrap up PCA

Week 12: 4/15 - 4/19

  • This week is an introduction to deep learning: Slides