STAT 432

Basics of Statistical Learning (with R and Python)

This is a newly designed version of STAT 432, which will replace the older version.

Read full course design notes

Previously, this course was designed to focus on mathematical formulation and implementing methods by hand to learn optimization directly. With the availability of AI tools, many programming tasks become less central. While you still need to understand concepts and code when needed, the focus now shifts toward understanding model properties and making sound choices in real analytics problems.

We will transition to a lecture/discussion format:

  • We introduce a new mathematical concept and model formulation on Tuesday.
  • All students submit a discussion question by Wednesday night for Thursday’s in-class discussion.
  • These questions will also appear on exams.
  • Most coding will be assisted by AI, while you are still encouraged to code on your own for homework. Homework questions will not be graded.
  • Feel free to use either R or Python in this course. Both versions of course materials will be provided.

This website is built with AI almost entirely.

Start with materials Read syllabus

Course Snapshot

Language

R and Python examples are presented side-by-side where possible.

Structure

Weekly lecture notes, homework guidance, and code demos in one place.

Delivery

Quarto site rendered to GitHub Pages via GitHub Actions.

Quick Start

  1. Read the syllabus for policy and grading details.
  2. Open materials.qmd for your first R/Python examples.
  3. Watch announcements on this page for updates.

Tentative Weekly Plan (Skeleton)

Week Topic Materials
1 Introduction and setup Add links to starter notebooks/slides
2 Linear models and evaluation Add R and Python demos
3 Regularization Add assignment and reading
4 Classification basics Add lab materials
5 Nonparametric methods Add section page
6+ Continue by topic Keep extending this table

Announcements

  • Site migrated to Quarto skeleton for 2026.
  • R and Python support now built into template pages.
  • Deployment configured with GitHub Actions to publish from GitHub.