Arrythmia-Classifier-using-GNN

Prerequisite

• pytorch (1.9.0+)
• dgl (0.8.0 / cuda 11.3)
• dgl installation = here
• networkx
• numpy
• torchmetrics

How to execute

1. put data into training and testing folder (data format : npz file)
2. run main.py

Notice

• The features used MIT-BIH arrhythmia dataset, it is refined in the BCML-KW. (Sharing is difficult.)
• Instead, there is an example code (breast cancer).