Memory-Based Dual Gaussian Processes for Sequential Learning

TL;DR

This paper introduces a memory-based dual Gaussian process method that enhances sequential learning accuracy by actively managing past data and errors, applicable to Bayesian optimization, active learning, and continual learning.

Contribution

It proposes a novel dual sparse variational Gaussian process approach that maintains and updates a memory of past data to improve sequential learning accuracy.

Findings

Effective in Bayesian optimization tasks

Improves active learning performance

Enhances continual learning accuracy

Abstract

Sequential learning with Gaussian processes (GPs) is challenging when access to past data is limited, for example, in continual and active learning. In such cases, errors can accumulate over time due to inaccuracies in the posterior, hyperparameters, and inducing points, making accurate learning challenging. Here, we present a method to keep all such errors in check using the recently proposed dual sparse variational GP. Our method enables accurate inference for generic likelihoods and improves learning by actively building and updating a memory of past data. We demonstrate its effectiveness in several applications involving Bayesian optimization, active learning, and continual learning.