Awesome

HAGEN: Homophily-Aware Graph Convolutional Recurrent Network for Crime Forecasting

This is a PyTorch implementation of HAGEN: Homophily-Aware Graph Convolutional Recurrent Network for Crime Forecasting.

Contact Update

Feel free to contact us if you have any questions related to our work:

Chenyu Wang: chenyuwang.monica@gmail.com Zongyu Lin: lzyxx17@gmail.com

Requirements

• scipy>=0.19.0
• numpy>=1.12.1
• pandas>=0.19.2
• pyyaml
• statsmodels
• torch
• tables
• future
• sklearn

Dependency can be installed using the following command:

pip install -r requirements.txt

Model Training

Here are commands for training the model on LA.

python hagen_train.py --config_filename ./crime-data/CRIME-LA/la_crime_9.yaml --month 9

Experimental settings and some supplemental results can be referred to HAGEN_suppl.pdf.

Citation

Please cite us if it is useful in your work:

@inproceedings{wang2022hagen,
  title={Hagen: Homophily-aware graph convolutional recurrent network for crime forecasting},
  author={Wang, Chenyu and Lin, Zongyu and Yang, Xiaochen and Sun, Jiao and Yue, Mingxuan and Shahabi, Cyrus},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={36},
  number={4},
  pages={4193--4200},
  year={2022}
}