This project aims to classify handwritten letters from the alphabet (A-Z) using a Multilayer Perceptron (MLP) neural network and Orthogonal Bipolar Vectors (OBVs) as target representations. The project involves training a neural network on labeled image data and testing its performance.