Gaussian Processes for Big Data

TL;DR

This paper presents a scalable variational inference method for Gaussian processes, enabling their application to large datasets with millions of points, and demonstrates its effectiveness on real-world data.

Contribution

It introduces stochastic variational inference for GPs, allowing efficient modeling of massive datasets and extension to non-Gaussian likelihoods and latent variable models.

Findings

Successfully applied to datasets with millions of points

Extended to non-Gaussian likelihoods and latent variable models

Demonstrated effectiveness on real-world data sets

Abstract

We introduce stochastic variational inference for Gaussian process models. This enables the application of Gaussian process (GP) models to data sets containing millions of data points. We show how GPs can be vari- ationally decomposed to depend on a set of globally relevant inducing variables which factorize the model in the necessary manner to perform variational inference. Our ap- proach is readily extended to models with non-Gaussian likelihoods and latent variable models based around Gaussian processes. We demonstrate the approach on a simple toy problem and two real world data sets.