25SummerStudy(25.07.~25.08.)

Study repository for foundation courses:
Linear Algebra, Probability & Information Theory, Machine Learning & Deep Learning, and Reinforcement Learning.

---

📘 Linear Algebra

• Boostcourse: Linear Algebra for AI

📊 Probability & Information Theory

• CS229 (Stanford) — Machine Learning (Summer 2019 Syllabus)
  YouTube Lectures-2019

🤖 Machine Learning & Deep Learning

• CS182 / 282A (UC Berkeley) — Designing, Visualizing and Understanding Deep Neural Networks
• CS294-158 (UC Berkeley, SP24) — Deep Unsupervised Learning
  YouTube Lectures-2021
• CS231n (Stanford) — Deep Learning for Computer Vision
  YouTube Lectures-2017
  YouTube Lectures-2025

🎮 Reinforcement Learning

• CS285 (UC Berkeley) — Deep Reinforcement Learning

---

📂 Repository Structure

25SummerStudy/
├─ cs231n/
│ ├─ python_tutorial/ # Python & NumPy tutorial
│ ├─ assignment1/ # 2025 (Original skeleton)
│ ├─ assignment1_colab/ # ✅ COMPLETED - k-NN, Softmax, Two-layer NN, Features, FC Networks
│ ├─ assignment2/ # 2025 (Original skeleton)
│ ├─ assignment2_colab/ # ✅ COMPLETED - BatchNorm, ConvNets, Dropout, PyTorch, RNN Captioning
│ ├─ assignment3/ # 2025 (Original skeleton)
│ └─ assignment3_colab/ # ✅ COMPLETED - Self-supervised Learning, Transformers, DDPM, CLIP/DINO
└─ cs182/
   ├─ cs182_hw1/ # 2022
   ├─ cs182_hw2/ # 2022
   └─ cs182_hw3/ # 2022

🎯 Completed Assignments

CS231n Assignment 1 ✅

• k-Nearest Neighbor (kNN): Distance calculations, cross-validation
• Softmax Classifier: Loss function and gradient implementation
• Two-layer Neural Network: Forward/backward pass, training
• Image Features: HOG, HSV color histograms
• Fully Connected Networks: Multi-layer networks with dropout/batch norm

CS231n Assignment 2 ✅

• Batch Normalization: Forward/backward pass implementation
• Convolutional Networks: Conv layers, pooling, spatial batch norm
• Dropout: Inverted dropout implementation
• PyTorch: Modern deep learning framework usage
• RNN Captioning: Image captioning with recurrent networks

CS231n Assignment 3 ✅

• Self-Supervised Learning: SimCLR contrastive learning
• Transformer Captioning: Attention mechanisms for image captioning
• DDPM: Denoising Diffusion Probabilistic Models
• CLIP/DINO: Vision-language models and self-supervised vision

📚 Sources

• CS231n materials: https://github.com/cs231n/cs231n.github.io
• CS182 SP22 materials: https://github.com/cs182sp22

This repository is for personal study only.
All copyrights belong to the respective course staffs and authors.
This repo contains only my personal notes and exercise solutions.