GLASSES: Relieving The Myopia Of Bayesian Optimisation

Javier GonzalezMichael OsborneNeil D. Lawrence
,  51:790-799, 2016.

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.

Cite this Paper


BibTeX
@InProceedings{pmlr-v-gonzalez-glasses16, title = {{GLASSES}: Relieving The Myopia Of {B}ayesian Optimisation}, author = {Javier Gonzalez and Michael Osborne and Neil D. Lawrence}, pages = {790--799}, year = {}, editor = {}, volume = {51}, address = {Cadiz, Spain}, url = {http://inverseprobability.com/publications/gonzalez-glasses16.html}, 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.} }
Endnote
%0 Conference Paper %T GLASSES: Relieving The Myopia Of Bayesian Optimisation %A Javier Gonzalez %A Michael Osborne %A Neil D. Lawrence %B %C Proceedings of Machine Learning Research %D %E %F pmlr-v-gonzalez-glasses16 %I PMLR %J Proceedings of Machine Learning Research %P 790--799 %U http://inverseprobability.com %V %W PMLR %X 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.
RIS
TY - CPAPER TI - GLASSES: Relieving The Myopia Of Bayesian Optimisation AU - Javier Gonzalez AU - Michael Osborne AU - Neil D. Lawrence BT - PY - DA - ED - ID - pmlr-v-gonzalez-glasses16 PB - PMLR SP - 790 DP - PMLR EP - 799 L1 - UR - http://inverseprobability.com/publications/gonzalez-glasses16.html AB - 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. ER -
APA
Gonzalez, J., Osborne, M. & Lawrence, N.D.. (). GLASSES: Relieving The Myopia Of Bayesian Optimisation. , in PMLR :790-799

Related Material