STGCN_COVID19_prediction

This repository is the term project of the course CSIT5910-HKUST.

Main contributions

(a) Learn a graph structure given the epidemic, population and geographical information.

(b) Use the STGCN to predict the confrimed cases.

(c) Visualization and comparison of the model with different paramters.

Declarations

The reference studies:

• J. Gao, R. Sharma, C. Qian, et.al. STAN: Spatio-Temporal Attention Network for Pandemic Prediction Using Real World Evidence. Jul. 2020, arXiv:2008.04215v1.

• B. Yu, H. Yin, Z. Zhu. Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting. Jul. 2018, arXiv:1709.04875v4.

The reference github repositories:

• https://github.com/v1xerunt/STAN
• https://github.com/FelixOpolka/STGCN-PyTorch

Other datases sourse:

• US epidemic information dataset.

Thanks for the contributions above!

Requirements

• Python 3.8
• PyTorch
• NumPy
• pickle
• matplotlib
• dgl