Implementation-of-Graph-Neural-Network-in-Python-on-MNIST-Dataset

Implementation of Graph Neural Network in Python on MNIST Dataset

Task I – Graph Neural Network Design In the graph learning framework you have set up, design one convolutional neural network and two different graph neural networks for node classification. (1) A convolutional neural network to classify the input images; the trained neural network will be used by the correct and smooth technique. (2) A graph network for transductive node classification using one or more spectral-based graph filters; your network must have at least two layers for graph filtering. (3) A graph network for transductive node classification using one or more spatial-based graph filters; your network must have at least two layers for graph filtering

Task II - Node Classification Techniques and Analyses You need to complete the following tasks and perform the required analyses. (1) DL Method - Train your convolutional neural network using the samples in the training set and then use the model to classify the samples in the test set. You need to report the loss and accuracy on the training and test set with respect to the number of training epochs and the confusion matrix for the test set of your final model