qEUBO: A Decision-Theoretic Acquisition Function for Preferential Bayesian Optimization

Raul Astudillo, Zhiyuan Jerry Lin, Eytan Bakshy, Peter Frazier
Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:1093-1114, 2023.

Abstract

Preferential Bayesian optimization (PBO) is a framework for optimizing a decision maker’s latent utility function using preference feedback. This work introduces the expected utility of the best option (qEUBO) as a novel acquisition function for PBO. When the decision maker’s responses are noise-free, we show that qEUBO is one-step Bayes optimal and thus equivalent to the popular knowledge gradient acquisition function. We also show that qEUBO enjoys an additive constant approximation guarantee to the one-step Bayes-optimal policy when the decision maker’s responses are corrupted by noise. We provide an extensive evaluation of qEUBO and demonstrate that it outperforms the state-of-the-art acquisition functions for PBO across many settings. Finally, we show that, under sufficient regularity conditions, qEUBO’s Bayesian simple regret converges to zero at a rate $o(1/n)$ as the number of queries, $n$, goes to infinity. In contrast, we show that simple regret under qEI, a popular acquisition function for standard BO often used for PBO, can fail to converge to zero. Enjoying superior performance, simple computation, and a grounded decision-theoretic justification, qEUBO is a promising acquisition function for PBO.

Cite this Paper


BibTeX
@InProceedings{pmlr-v206-astudillo23a, title = {qEUBO: A Decision-Theoretic Acquisition Function for Preferential Bayesian Optimization}, author = {Astudillo, Raul and Lin, Zhiyuan Jerry and Bakshy, Eytan and Frazier, Peter}, booktitle = {Proceedings of The 26th International Conference on Artificial Intelligence and Statistics}, pages = {1093--1114}, year = {2023}, editor = {Ruiz, Francisco and Dy, Jennifer and van de Meent, Jan-Willem}, volume = {206}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v206/astudillo23a/astudillo23a.pdf}, url = {https://proceedings.mlr.press/v206/astudillo23a.html}, abstract = {Preferential Bayesian optimization (PBO) is a framework for optimizing a decision maker’s latent utility function using preference feedback. This work introduces the expected utility of the best option (qEUBO) as a novel acquisition function for PBO. When the decision maker’s responses are noise-free, we show that qEUBO is one-step Bayes optimal and thus equivalent to the popular knowledge gradient acquisition function. We also show that qEUBO enjoys an additive constant approximation guarantee to the one-step Bayes-optimal policy when the decision maker’s responses are corrupted by noise. We provide an extensive evaluation of qEUBO and demonstrate that it outperforms the state-of-the-art acquisition functions for PBO across many settings. Finally, we show that, under sufficient regularity conditions, qEUBO’s Bayesian simple regret converges to zero at a rate $o(1/n)$ as the number of queries, $n$, goes to infinity. In contrast, we show that simple regret under qEI, a popular acquisition function for standard BO often used for PBO, can fail to converge to zero. Enjoying superior performance, simple computation, and a grounded decision-theoretic justification, qEUBO is a promising acquisition function for PBO.} }
Endnote
%0 Conference Paper %T qEUBO: A Decision-Theoretic Acquisition Function for Preferential Bayesian Optimization %A Raul Astudillo %A Zhiyuan Jerry Lin %A Eytan Bakshy %A Peter Frazier %B Proceedings of The 26th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2023 %E Francisco Ruiz %E Jennifer Dy %E Jan-Willem van de Meent %F pmlr-v206-astudillo23a %I PMLR %P 1093--1114 %U https://proceedings.mlr.press/v206/astudillo23a.html %V 206 %X Preferential Bayesian optimization (PBO) is a framework for optimizing a decision maker’s latent utility function using preference feedback. This work introduces the expected utility of the best option (qEUBO) as a novel acquisition function for PBO. When the decision maker’s responses are noise-free, we show that qEUBO is one-step Bayes optimal and thus equivalent to the popular knowledge gradient acquisition function. We also show that qEUBO enjoys an additive constant approximation guarantee to the one-step Bayes-optimal policy when the decision maker’s responses are corrupted by noise. We provide an extensive evaluation of qEUBO and demonstrate that it outperforms the state-of-the-art acquisition functions for PBO across many settings. Finally, we show that, under sufficient regularity conditions, qEUBO’s Bayesian simple regret converges to zero at a rate $o(1/n)$ as the number of queries, $n$, goes to infinity. In contrast, we show that simple regret under qEI, a popular acquisition function for standard BO often used for PBO, can fail to converge to zero. Enjoying superior performance, simple computation, and a grounded decision-theoretic justification, qEUBO is a promising acquisition function for PBO.
APA
Astudillo, R., Lin, Z.J., Bakshy, E. & Frazier, P.. (2023). qEUBO: A Decision-Theoretic Acquisition Function for Preferential Bayesian Optimization. Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 206:1093-1114 Available from https://proceedings.mlr.press/v206/astudillo23a.html.

Related Material