Handwritten-Digit-MLP-Classification Using the MNIST database. It is a collection of 70000 handwritten digits split into training and test set of 60000 and 10000 images respectively. The model is built using Multi Layer Perceptron (MLP).