Classification of FashionMNIST dataset using CNN

Machine Learning for Robotics II - Final Project

• Syed Hasan Shozab Abidi, 4929631@studenti.unige.it
• Muhammad Ali Haider Dar, 5046263@studenti.unige.it

MSc Robotics Engineering, University of Genoa, Italy

Instructor: Prof. Luca Oneto

Overview

The aim of this project is to perform multi-class classification on MNIST Fashion dataset using Convolution Neural Network (CNN). The neural network is build from scratch using PyTorch library. PyTorch is an environment for programming neural networks with the encapsulated effect that you can use hardware accelerators, such as CUDA. It is built on top of TensorFlow which is used for distributed machine learning on any platform.

The Jupyter notebook in this repository consists of the code and comments for the project.