Predicting the temporal and geographical occurrence of conflicts in Myanmar with two paradigms of spatiotemporal networks.
The file 'Spatiotemporal graph prediction.ipynb' contains instructions for usage and reproduction of results. For an overview of the project's background and results, see the slides below.
We compared two models from recent papers, each instantiating a different paradigm in dynamic graph prediction. The first model was from the following paper:
"Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting"
Lei Bai, Lina Yao, Can Li, Xianzhi Wang, Can Wang
The second was from this paper:
"On the Equivalence Between Temporal and Static Graph Representations for Observational Predictions"
Jianfei Gao, Bruno Ribeiro
Created by Mac O'Brien and Hset Hset Naing
