Update May 30: see https://github.com/thangbui/geepee for a faster implementation, which also has a lot more things than just deep GPs. There are some DGP examples for regression and binary classification here https://github.com/thangbui/geepee/blob/master/examples/dgpr_aep_examples.py

Update Mar 27: New version coming soon!

This repository contains Theano and python code for the paper:

Deep Gaussian Processes for Regression using Approximate Expectation Propagation, Thang Bui, Daniel Hernandez-Lobato, Yingzhen Li, Jose miguel Hernandez-Lobato and Richard Turner, ICML 2016.

Please see this Theano example and this pure Python example for usage.