Deep Gaussian Processes with Importance-Weighted Variational Inference

back to our research

Deep Gaussian Processes with Importance-Weighted Variational Inference

June 9 - 15 2019 ICML, Long Beach, USA

Authors: Hugh Salimbeni, Vincent Dutordoir, James Hensman and Marc Deisenroth (PROWLER.io
and Imperial College London)

Abstract: Deep Gaussian processes (DGPs) can model complex marginal densities as well as complex mappings. Non-Gaussian marginals are essential for modelling real-world data, and can be generated from the DGP by incorporating uncorrelated variables to the model. Previous work in the DGP model has introduced noise additively and used variational inference with a combination of sparse Gaussian processes and mean-field Gaussians for the approximate posterior. Additive noise attenuates the signal, and the Gaussian form of variational distribution may lead to an inaccurate posterior. We instead incorporate noisy variables as latent covariates, and propose a novel importance-weighted objective, which leverages analytic results and provides a mechanism to trade off computation for improved accuracy. Our results demonstrate that the importance-weighted objective works well in practice and consistently outperforms classical variational inference, especially for deeper models.

Probabilistic Modelling

Gaussian Processes


See paper

Join us to make AI that will change the world

join our team