Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SMASH: Score Matching-based Pseudolikelihood Estimation of Neural Marked Spatio-Temporal Point Process

This code is for SMASH main experiment on three marked STPP datasets: Earthquake, Crime and Football. Data files are included. The code is tested under a Linux desktop with torch 2.0 and Python 3.11.4.

Environment

  • Tested OS: Linux
  • Python >= 3.7
  • PyTorch == 2.0.1

Model Training

Use the following command to train and test SMASH on Earthquake dataset:

cd ./scripts bash train_earthquake.sh

Note

The implemention is based on one of the baselines: DSTPP.

About

Code for Paper "Beyond Point Prediction: Score Matching-based Pseudolikelihood Estimation of Neural Marked Spatio-Temporal Point Process"

Resources

Stars

6 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages