Skip to content

sandeepsoni/MHP

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 

Repository files navigation

MHP

This repository contains a light-weight python implementation for generating events and estimating the parameters using a Multivariate Hawkes Process. It contains the following code components:

  • HP/simulators: To generate a cascade with given parameters using the modified Ogata's thinning algorithm.
  • HP/estimators: To estimate the parameters of multivariate hawkes process using maximum likelihood.

The simulation code is copied and modified slightly from Steve Morse's excellent implementation. The code is optimized to get runtime improvements compared to naive implementations.

If you happen to use this code, consider citing our paper. Here's the open-access version

@inproceedings{soni2019detecting,
  title={Detecting Social Influence in Event Cascades by Comparing Discriminative Rankers},
  author={Soni, Sandeep and Ramirez, Shawn Ling and Eisenstein, Jacob},
  booktitle={The 2019 ACM SIGKDD Workshop on Causal Discovery},
  pages={78--99},
  year={2019}
}

If you discover any bug, please file an issue. For contributions, please make a pull request.

TODO

  • Add Jupyter notebook demostrating how to use both the generation and estimation code

Other similar resources

About

Generation and MLE Estimation for Multivariate Hawkes Process

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages