https://github.com/shchur/triangular-tpp

Implementation of "Fast and Flexible Temporal Point Processes with Triangular Maps" (Oral @ NeurIPS 2020)
https://github.com/shchur/triangular-tpp

normalizing-flows pytorch temporal-point-processes

Last synced: about 1 month ago
JSON representation

Implementation of "Fast and Flexible Temporal Point Processes with Triangular Maps" (Oral @ NeurIPS 2020)

• Host: GitHub
• URL: https://github.com/shchur/triangular-tpp
• Owner: shchur
• License: mit
• Created: 2020-10-09T21:50:33.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2023-12-01T14:26:50.000Z (about 1 year ago)
• Last Synced: 2024-11-30T00:06:20.522Z (about 1 month ago)
• Topics: normalizing-flows, pytorch, temporal-point-processes
• Language: Jupyter Notebook
• Homepage: https://arxiv.org/abs/2006.12631
• Size: 9.23 MB
• Stars: 23
• Watchers: 2
• Forks: 7
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Fast and Flexible Temporal Point Processes with Triangular Maps



  



This repository includes a reference implementation of the algorithms described in ["Fast and Flexible Temporal Point Processes with Triangular Maps"](https://arxiv.org/abs/2006.12631) by Oleksandr Shchur, Nicholas Gao, Marin Biloš and Stephan Günnemann (Oral, NeurIPS 2020).

- [Video explaining the paper](https://videos.neurips.cc/video/slideslive-38938478)

Temporal point processes (TPPs) allow us to define probability distributions over variable-length event sequences in some time interval `[0, t_max]`.

In our paper, we show how to define TPPs using invertible transformations, similar to normalizing flows.

The code includes new parametrizations for several existing TPPs as well as a new, more flexible model.

Our parametrizations allow to both draw samples & compute likelihood in parallel, which leads to signficant speedups compared to traditional RNN-based models.

The following models are available in [`ttpp.models`](https://github.com/shchur/triangular-tpp/blob/main/ttpp/models.py):

- `Inhomogeneous Poisson Process`

- `Renewal Process`

- `Modulated Renewal Process`

- `TriTPP`

- `Autoregressive` (RNN-based TPP with slow sampling)

## Requirements

The code is written in Python version 3.7 and was tested on Ubuntu 18.04.

The code requires PyTorch version 1.5 with CUDA enabled.

Other requirements are listed in [`requirements.txt`](requirements.txt).

To install the library run

```

pip install -e .

```

The datasets used in the paper can be found in `data/`.

## Usage

Jupyter notebooks reproducing the experimental results can be found in the `notebooks/` folder:

- [Differentiable relaxation](notebooks/differentiable_relaxation.ipynb)

- [Scalability](notebooks/scalability.ipynb)

- [Density estimation](notebooks/density_estimation.ipynb)

- [Variational inference for MJPs](notebooks/variational_inference.ipynb)

You can also train the model using command line. For example, to train the `TriTPP` model on the `taxi` dataset run

```

python scripts/experiment.py taxi TriTPP

```

To see the command line arguments, use

```

python scripts/experiment.py --help

```

## Cite

Please cite our paper if you use this code or data in your own work:

```

@inproceedings{shchur2020fast,

  title = {Fast and Flexible Temporal Point Processes with Triangular Maps},

  author = {Shchur, Oleksandr and Gao, Nicholas and Bilo\v{s}, Marin and G{\"u}nnemann, Stephan},

  booktitle={Advances in Neural Information Processing Systems (NeurIPS)},

  year = {2020} 

}

```