Timezone: »
Warped Gaussian processes (WGP) [1] model output observations in regression tasks as a parametric nonlinear transformation of a Gaussian process (GP). The use of this nonlinear transformation, which is included as part of the probabilistic model, was shown to enhance performance by providing a better prior model on several data sets. In order to learn its parameters, maximum likelihood was used. In this work we show that it is possible to use a non-parametric nonlinear transformation in WGP and variationally integrate it out. The resulting Bayesian WGP is then able to work in scenarios in which the maximum likelihood WGP failed: Low data regime, data with censored values, classification, etc. We demonstrate the superior performance of Bayesian warped GPs on several real data sets.
Author Information
Miguel Lazaro-Gredilla (Vicarious AI)
More from the Same Authors
-
2021 Poster: Perturb-and-max-product: Sampling and learning in discrete energy-based models »
Miguel Lazaro-Gredilla · Antoine Dedieu · Dileep George -
2013 Poster: Variational Inference for Mahalanobis Distance Metrics in Gaussian Process Regression »
Michalis Titsias · Miguel Lazaro-Gredilla -
2011 Poster: Spike and Slab Variational Inference for Multi-Task and Multiple Kernel Learning »
Michalis Titsias · Miguel Lazaro-Gredilla -
2009 Poster: Inter-domain Gaussian Processes for Sparse Inference using Inducing Features »
Miguel Lazaro-Gredilla · Anibal R Figueiras-Vidal