Graph Convolutional Network on MNIST Handwritten digits dataset

In this repository Graph Convolutional Network and Convolutional Neural Network is implemented using pytorch and pytorch geometric. \

Dataset: MNIST Handwritten digits dataset.

For more details refer to: https://github.com/KenaHemnani/gcn-image-classification-task/blob/master/Report.pdf In the above pdf GCN is explained in simple language.

Problem Formulation: Image classification is formulated as a Node Classification Task.

Clone the repository:

git clone https://github.com/KenaHemnani/gcn-image-classification-task.git

Create and activate the python virtual environment:

python3 -m venv gcn_env
source gcn_env/bin/activate

Intsall requirements:

pip install -r requirements.txt

Run jupyter:

jupyter notebook

Now, Open the CNN_Mnist.ipynb and GCN_Mnist.ipynb file execute the code.