Classification Project

Project Links

• Documentation & Report: https://thomaschiari.github.io/deep-learning-coursework/classification-project/classification-project/
• Source Code: https://github.com/thomaschiari/deep-learning-classification-project

Overview

This project implements a deep learning classification system using a Multi-Layer Perceptron (MLP) built from scratch with NumPy. The goal is to predict whether a client will subscribe to a term deposit based on the Bank Marketing Dataset.

Key Features

• From-scratch MLP implementation using only NumPy
• Early stopping and regularization techniques
• Comprehensive evaluation with ROC-AUC and PR-AUC metrics
• Detailed analysis of training dynamics and overfitting behavior
• Complete documentation with step-by-step implementation guide

For detailed implementation, analysis, and results, please visit the links above.