Skip to content

Python implementation of the paper "Modeling Extreme Events in Time Series Prediction", with some modifications.

Notifications You must be signed in to change notification settings

HuiqunHuang/EVL

Repository files navigation

EVL

Introduction

This repo is the personal Python implementation of paper "Modeling Extreme Events in Time Series Prediction", [paper], with some modifications, not official codes.

Environment and Dependencies

  • Python 3.6
  • Tensorflow-GPU-2.3.0
  • Keras 2.7.0
  • Pandas 1.1.5
  • Scikit-learn 0.23.1
  • CUDA 10.1
  • CuDNN 7.6

Model Training & Evaluation

python MainPredictionFunction/Chicago_EVL_Main.py

Citations

If you were using our codes or found this repository useful, please consider citing our work:

@article{huang2021multi,
  title={Multi-Head Spatio-Temporal Attention Mechanism for Urban Anomaly Event Prediction},
  author={Huang, Huiqun and Yang, Xi and He, Suining},
  journal={Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies},
  volume={5},
  number={3},
  pages={1--21},
  year={2021},
  publisher={ACM New York, NY, USA}
}
@inproceedings{huang2023extreme,
  title={Extreme-Aware Local-Global Attention for Spatio-Temporal Urban Mobility Learning},
  author={Huang, Huiqun and He, Suining and Tabatabaie, Mahan},
  booktitle={2023 IEEE 39th International Conference on Data Engineering (ICDE)},
  pages={1059--1070},
  year={2023},
  organization={IEEE}
}

About

Python implementation of the paper "Modeling Extreme Events in Time Series Prediction", with some modifications.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published