https://github.com/alan-turing-institute/rpsmf
Code for Probabilistic Sequential Matrix Factorization
https://github.com/alan-turing-institute/rpsmf
bayesian-methods matrix-factorization online-algorithms probabilistic-models reproducible-research reproducible-science sequential-data time-series time-series-analysis time-series-forecasting
Last synced: about 1 month ago
JSON representation
Code for Probabilistic Sequential Matrix Factorization
- Host: GitHub
- URL: https://github.com/alan-turing-institute/rpsmf
- Owner: alan-turing-institute
- License: mit
- Created: 2020-09-04T12:29:50.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2021-04-27T20:41:53.000Z (about 4 years ago)
- Last Synced: 2025-04-09T01:51:20.982Z (2 months ago)
- Topics: bayesian-methods, matrix-factorization, online-algorithms, probabilistic-models, reproducible-research, reproducible-science, sequential-data, time-series, time-series-analysis, time-series-forecasting
- Language: Python
- Homepage: http://proceedings.mlr.press/v130/akyildiz21a.html
- Size: 20.5 MB
- Stars: 15
- Watchers: 6
- Forks: 5
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Probabilistic Sequential Matrix Factorization
This repository contains the code to reproduce the experiments in:
[**Akyildiz, van den Burg, Damoulas, Steel - Probabilistic Sequential Matrix
Factorization (2021)**](http://proceedings.mlr.press/v130/akyildiz21a.html)
Work that uses the methods described in the paper or the code in this
repository should cite the paper, for instance using the following BibTeX
entry:
```bib
@inproceedings{akyildiz2021probabilistic,
title={Probabilistic Sequential Matrix Factorization},
author={{\"O}mer Deniz Akyildiz and Gerrit J. J. {van den Burg} and Theodoros Damoulas and Mark F. J. Steel},
booktitle={Proceedings of the 24th International Conference on Artificial Intelligence and Statistics},
editor={Banerjee, Arindam and Fukumizu, Kenji},
pages={3484--3492},
volume={130},
year={2021},
series={Proceedings of Machine Learning Research},
publisher={PMLR},
}
```
If you encounter a problem when using this repository or simply want to ask a
question, please don't hesitate to [open an issue on
GitHub](https://github.com/alan-turing-institute/rPSMF) or send an email to
``odakyildiz at turing dot ac dot uk`` and/or ``gertjanvandenburg at gmail dot
com``.
## Introduction
Our Probabilistic Sequential Matrix Factorization (PSMF) method allows you to
model high-dimensional timeseries data that exhibits non-stationary dynamics.
We also propose a robust variant of the model, called rPSMF, that handles
model misspecification and outliers.
See [the paper](http://proceedings.mlr.press/v130/akyildiz21a.html) for
further details.
Illustration of using rPSMF to model a 20-dimensional time series
with non-linear dynamics and t-distributed outliers. Blue lines are the
observed data and yellow lines are unobserved future data. The red dashed
line shows the predictions from our model.
## Usage
The code in this repository is organized by experiment. The
[Convergence](./Convergence) directory contains the code for the convergence
experiment in the appendix of the paper.
The accompanying Makefile can be used to reproduce the experiments by simply
running
```bash
$ make ExperimentSynthetic # or ExperimentBeijing, or ExperimentImpute, or ExperimentRecursive
```
These experiments will be run through a Python virtual environment that will
be automatically created with the required dependencies. For the imputation
experiment the results are captured in LaTeX tables that are automatically
generated as well.
The code for the change point detection Experiment and the convergence
experiment are written in Matlab, so require a Matlab installation to
reproduce. See the readme files in the corresponding directories for more
information.
## Notes
The code is licensed under the MIT license unless otherwise noted, see the
[LICENSE](./LICENSE) file for further details.