This project implements neural network and convolutional neural network. The task is to carry out classification on Fashion-MNIST dataset. There are 10 classes of different types of clothing. Our task is to recognize an image and identify it as one of the ten classes. We will train classifiers using images from Zalando’s clothing article. The project is divided into three tasks. The first task is to build single layer(hidden) neural network from scratch in python. The second task would be to build a multi-layer Neural Network with open-source neural-network library, Keras. The third task would be to build Convolutional Neural Network (CNN) with open-source neural-network library, Keras. Layers in all three types of networks will be trained and tested on Fashion-MNIST dataset. The inference will consist of comparison between accuracy and loss between all three ways to classify and predict data using neural networks.