This repository contains a collection of my personal projects in machine learning, deep learning, and other areas of computer science.
A PyTorch-based MNIST digit classification project that achieves 99.4% validation accuracy with less than 20k parameters. Features include:
- Convolutional Neural Network with Batch Normalization
- Dropout for regularization
- Fully Connected layers
- Training completes in less than 20 epochs
Personal_Projects/
├── README.md
└── MNIST_99.4/
├── models/
├── tests/
├── train.py
└── utils.py