This repository serves as a comprehensive technical portfolio for the Stanford CS229: Machine Learning (Summer Edition). It integrates theoretical derivations, numerical implementations, and a formal research project.
As an Electrical and Electronics Engineer specializing in AI/ML, I have specifically chosen the Summer 2020 Edition (instructed by Anand Avati) for its unique focus:
- Mathematical Rigor: Unlike standard sessions, this version emphasizes the analytical derivations of algorithms.
- First Principles: Focuses on foundational Matrix Calculus, Probability, and Optimization.
- Advanced Topics: Includes in-depth coverage of Gaussian Processes, EM variants, and Variational Autoencoders (VAE).
Course Link: Stanford CS229 Summer 2020
/Lecture Notes: Official CS229 Summer 2020 lecture PDFs (Lectures 01–12, Full Notes, Deep Learning Notes). Personal study notes maintained in Obsidian — not tracked in this repo./Problem Sets: Original problem set PDFs (PS0–PS3)./Problem Sets Solutions: Worked solutions (PS0 complete; PS1–PS3 in progress)./Final Project: CS229 capstone — Knowledge Distillation ablation study (Logit-KD vs Feature-KD, ResNet-50 → ResNet-18 on CIFAR-10).paper/— Full project report (PDF + LaTeX source)poster/— Project poster (PDF + LaTeX source)- Full implementation: github.com/umutonuryasar/kd-cifar10
- OS: Ubuntu
-
Notes: Obsidian (Markdown with
$\LaTeX$ rendering) - Environment: Python 3.x, NumPy, Matplotlib, Jupyter Lab
-
Documentation:
$\LaTeX$ (for the final paper and poster)
This project is licensed under the MIT License - see the LICENSE file for details.