Skip to content

Latest commit

 

History

History
10 lines (9 loc) · 807 Bytes

README.md

File metadata and controls

10 lines (9 loc) · 807 Bytes

Handwritten-digit image classification with Neural Network

This project deals with an introductory example in Deep Learning. A subset of hand-written digits data has been taken from Yann LeCun's website. The code does not deal much with data manipulation as the images have already been centered and appropriately transformed.

Libraries required.

The following libraries are required for the functioning of the code. PyTorch, torchvision, sklearn, numpy and matplotlib.

This project employs PyTorch to define the neural network and the optimizer. But first, numpy needs to be installed.

pip install numpy

For sklearn

pip install -U scikit-learn

The instructions to download pytorch for CPU can be obtained here.