GP-ALPS: Automatic Latent Process Selection for Multi-Output Gaussian Process Models

TL;DR

GP-ALPS introduces an automated method for selecting the number and kernels of latent processes in multi-output Gaussian processes, reducing manual tuning and bias, through a variational inference scheme.

Contribution

It presents a novel automatic latent process selection method for multi-output GPs using variational inference, improving model efficiency and objectivity.

Findings

Automatically selects relevant latent processes.

Demonstrates comparable performance to MCMC.

Effective in preliminary experiments.

Abstract

A simple and widely adopted approach to extend Gaussian processes (GPs) to multiple outputs is to model each output as a linear combination of a collection of shared, unobserved latent GPs. An issue with this approach is choosing the number of latent processes and their kernels. These choices are typically done manually, which can be time consuming and prone to human biases. We propose Gaussian Process Automatic Latent Process Selection (GP-ALPS), which automatically chooses the latent processes by turning off those that do not meaningfully contribute to explaining the data. We develop a variational inference scheme, assess the quality of the variational posterior by comparing it against the gold standard MCMC, and demonstrate the suitability of GP-ALPS in a set of preliminary experiments.