Skip to content
Deep Gaussian Processes with Doubly Stochastic Variational Inference
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
demos
doubly_stochastic_dgp added step up functionality Jun 29, 2018
media added presentation Oct 20, 2017
tests added step up functionality Jun 29, 2018
.gitignore Update .gitignore May 23, 2017
LICENSE Initial commit May 16, 2017
README.md state that gpflow version must be =1.1 Feb 21, 2019
setup.py added setup Jan 22, 2018

README.md

Doubly-Stochastic-DGP

Deep Gaussian Processes with Doubly Stochastic Variational Inference

Requirements: gpflow1.1.1 and tensorflow1.8. NB not compatabile with more recent versions (e.g. gpflow1.2)

This code accompanies the paper

@inproceedings{salimbeni2017doubly, title={Doubly stochastic variational inference for deep gaussian processes}, author={Salimbeni, Hugh and Deisenroth, Marc}, booktitle={Advances in Neural Information Processing Systems}, year={2017} }

See the arxiv version at https://arxiv.org/abs/1705.08933

This code now offers additional functionality than in the above paper. In particular, natural gradients are now supported. If you use these, please consider citing the following paper:

@inproceedings{salimbeni2018natural, title={Natural Gradients in Practice: Non-Conjugate Variational Inference in Gaussian Process Models}, author={Salimbeni, Hugh and Eleftheriadis, Stefanos and Hensman, James}, booktitle={Artificial Intelligence and Statistics}, year={2018} }

You can’t perform that action at this time.