Schedule 2026

Problem Sets and Exams

Clone the notebooks repository to get the Jupyter notebooks, and submit on Canvas.

1. Due Midnight PST on January 13 - Problem Set 1 and ipynb
2. Due Midnight PST on January 25 - Problem Set 2 and ipynb
3. Due Midnight PST on February 8 - Problem Set 3 and ipynb
4. Due Midnight PST on February 22 - Problem Set 4 and ipynb

• If you cannot get free API tokens for OpenAI then use Gemini, another AI model, or skip that question

Schedule

1. January 5 - Course Overview

2. January 7 - Representation learning in Course Overview and Python Frameworks

3. January 12 - Rethinking Least Squares and Iterative Methods (only conditioning, and preconditioning)

   • Self-study in Iterative Methods: Stationary Iterative Methods, Structured Linear Operators, Krylov Subspace Methods, and eigenvalue problems
   • Self-study in Factorizations and Direct Methods, especially Computational Complexity, Matrix Structure, and Matrix Factorizations

4. January 14 - Rethinking Least Squares and start Differentiable Everything!

5. January 19 - Finish Differentiable Everything!

6. January 21 - High-Dimensional Optimization for ML

7. January 26 - High-Dimensional Optimization for ML and Optimization in JAX

8. January 28 - Deep Learning and Representation Learning

9. February 2 - Overparameterization and Inductive Bias and High Dimensional Probability and Concentration of Measure

10. February 4 - Finish High Dimensional Probability and Concentration of Measure

11. February 9 - Kernel Methods and Gaussian Processes and Embeddings, NLPs, and LLMs

12. February 11 - Finish Embeddings, NLPs, and LLMs and Transformers and Attention

13. February 16 - SPRING BREAK (NO CLASS)

14. February 18 - SPRING BREAK (NO CLASS)

15. February 23 - Switch to second half of course