This repository contains various jupyter pages written by me working on the MNIST datasets for my course Pattern Recognition. It uses different learning methods such as Support Vector Machines, Neural Networks, Generative Models, Probabilistic Graphic Models and Linear Discriminant functions. It uses keras and tensorflow for most of the codes.

Implementation of Conditional Generative Adversarial Nets with PyTorch.

Implementing Hand Written Digit Recognition using MNIST Datasets with the help of Tensor-Flow. It does the classification with 99% of accuracy without using the fixed number of epochs. It stops training when required level of accuracy is achieved.

Implementation of digit recognition using the MNIST dataset. Train a model on grayscale handwritten digit images to accurately classify and predict numerical values of unseen digits.