Handwritten Digit Recognition (CNN)

Deep learning classifier trained on MNIST with a live prediction demo

A convolutional neural network that classifies handwritten digits, trained on the MNIST dataset with a simple web UI where you draw a digit and get a live prediction.

What's included

• CNN model trained and evaluated on MNIST
• Live drawable canvas for real-time prediction
• Model accuracy/loss plots included
• Well-documented training notebook
• Project report and viva Q&A notes

Tech stack