GLASSES: Relieving The Myopia Of Bayesian Optimisation

TL;DR

GLASSES is a novel global optimisation algorithm that uses look-ahead simulation and expected-loss search to consider dozens of future evaluations, significantly improving performance over traditional myopic methods.

Contribution

It introduces a scalable approach to non-myopic Bayesian optimisation by approximating the look-ahead loss with stochastic simulation and expectation propagation.

Findings

Substantive performance gains demonstrated in empirical tests.

Enables consideration of dozens of future evaluations.

Outperforms traditional myopic optimisation methods.

Abstract

We present GLASSES: Global optimisation with Look-Ahead through Stochastic Simulation and Expected-loss Search. The majority of global optimisation approaches in use are myopic, in only considering the impact of the next function value; the non-myopic approaches that do exist are able to consider only a handful of future evaluations. Our novel algorithm, GLASSES, permits the consideration of dozens of evaluations into the future. This is done by approximating the ideal look-ahead loss function, which is expensive to evaluate, by a cheaper alternative in which the future steps of the algorithm are simulated beforehand. An Expectation Propagation algorithm is used to compute the expected value of the loss.We show that the far-horizon planning thus enabled leads to substantive performance gains in empirical tests.