This repository presents a deep learning–based approach to medical image classification, developed as part of the Applied Machine Learning Systems (AMLS) coursework.
It includes modular CNN implementations for classifying images from the BreastMNIST and BloodMNIST datasets, with a structured design that separates task-specific experimentation from reusable model components. A central execution script (main.py) enables streamlined training and evaluation.
The project highlights practical experience in building, organizing, and applying deep learning pipelines to real-world medical datasets.