Image_Classification_ML

CSE382 Image Classifiers Project

This project implements a complete image-classification pipeline on the MNIST dataset, starting with binary classification and then extending to multi-class classification.

Project Objectives

Phase 1: Baseline System

Build a binary image-classification pipeline for two selected MNIST classes using at least three machine learning algorithms.

Phase 2: Improvement and Enhancement

Extend the system to all 10 MNIST classes and improve performance using advanced techniques such as feature engineering, hyperparameter tuning, regularization, ensemble methods, and model analysis.

Dataset

The project uses the MNIST handwritten digits dataset.