Skip to content
Source code of the neural Hawkes particle smoothing (ICML 2019)
Branch: master
Clone or download
Latest commit 8f33c75 May 16, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data add README May 14, 2019
mp_manager mp_manager, distances, and miss May 12, 2019
nhps mp_manager, distances, and miss May 12, 2019
.gitignore functions, io, models, sigtest May 12, 2019
LICENSE Initial commit May 11, 2019
README.md update paper link May 15, 2019
setup.py functions, io, models, sigtest May 12, 2019

README.md

The Neural Hawkes Particle Smoothing

Source code for Imputing Missing Events in Continuous-Time Event Streams (ICML 2019) runnable on GPU and CPU.

Reference

If you use this code as part of any published research, please acknowledge the following paper (it encourages researchers who publish their code!):

@inproceedings{mei-2019-smoothing,
  author =      {Hongyuan Mei and Guanghui Qin and Jason Eisner},
  title =       {Imputing Missing Events in Continuous-Time Event Streams},
  booktitle =   {Proceedings of the International Conference on Machine Learning},
  year =        {2019}
}

Instructions

Here are the instructions to use the code base.

Dependencies

This code is written in Python 3, and I recommend you to install:

  • Anaconda that provides all the Python-related dependencies;
  • PyTorch 1.0 that handles auto-differentiation.

Prepare Data

Download datasets from this Google Drive link to the 'data' folder. See more details in this README.

Install Modules

Run the command line below to install the modules (add -e option if you need an editable installation):

pip install .

Train Models

Go to the nhps/functions directory.

To train the neural Hawkes process with complete data, try the command line below for detailed guide (see section 2 in paper for more technical details):

python train_nhpf.py --help

To train the neural Hawkes particle smoother with incomplete data, try the command line below for detailed guide (see section 3 in paper for more technical details):

python train_nhps.py --help

Test Models

Go to the nhps/functions directory.

To evaluate (dev or test), use the command line below for detailed guide (see section 4 in paper for more technical details):

python test.py --help

License

This project is licensed under the BSD 3-Clause License - see the LICENSE file for details.

You can’t perform that action at this time.