My homework for CSE 446/546: Machine Learning
at the University of Washington, Autumn 2024 (graduate section). Each hw folder has my code for that assignment's programming
problems, and the PDFs are my written solutions.
The problem statements, starter code and tests came from the course staff. The implementations are mine.
| Assignment | Programming problems |
|---|---|
| hw0 | the central limit theorem with empirical CDFs, plain Python vs NumPy |
| hw1 | polynomial regression and learning curves, a ridge regression classifier on MNIST |
| hw2 | lasso with ISTA on crime data, binary logistic regression |
| hw3 | PyTorch building blocks from scratch (layers, losses, SGD), polynomial and RBF kernels with cross-validation and the bootstrap, neural networks on MNIST |
| hw4 | k-means clustering on MNIST |
Each problem folder has its own README with the original instructions.
I use a lot of this in my basketball analytics work: wnba-analysis.