Continual Gaussian Processes

This repository contains the implementation of our Continual (Multi-task) Gaussian Process model. We provide a detailed code for single-output GP regression, multi-output GP regression, GP classification and long-term continual learning.

Please, if you use this code, cite the following preprint:

@article{MorenoArtesAlvarez19,
  title = {Continual Multi-task Gaussian Processes},
  author = {Moreno-Mu\~noz, Pablo and Art\'es-Rodr\'iguez, Antonio and \'Alvarez, Mauricio A},
  journal = {arXiv preprint arXiv:1911.00002},
  year = {2019}
}

Solar sunspots data.

Results: In the /experiments/ folder you may find the following scripts for simulations.

single_output.py // Continual GP regression
multi_output.py  // Continual multi-output GP regression
banana.py        // Continual GP classification
solar.py         // Long-term continual GP regression (figure above).

The Python syntaxes of likelihood distributions and the structure of our code is based on the HetMOGP repository.

Contributors

Pablo Moreno-Muñoz, Antonio Artés-Rodríguez and Mauricio A. Álvarez

For further information or contact:

[email protected]