CO-BED: Information-Theoretic Contextual Optimization via Bayesian Experimental Design

Desi R. Ivanova, Joel Jennings, Tom Rainforth, Cheng Zhang, Adam Foster
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:14445-14464, 2023.

Abstract

We formalize the problem of contextual optimization through the lens of Bayesian experimental design and propose CO-BED—a general, model-agnostic framework for designing contextual experiments using information-theoretic principles. After formulating a suitable information-based objective, we employ black-box variational methods to simultaneously estimate it and optimize the designs in a single stochastic gradient scheme. In addition, to accommodate discrete actions within our framework, we propose leveraging continuous relaxation schemes, which can naturally be integrated into our variational objective. As a result, CO-BED provides a general and automated solution to a wide range of contextual optimization problems. We illustrate its effectiveness in a number of experiments, where CO-BED demonstrates competitive performance even when compared to bespoke, model-specific alternatives.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-ivanova23a, title = {{CO}-{BED}: Information-Theoretic Contextual Optimization via {B}ayesian Experimental Design}, author = {Ivanova, Desi R. and Jennings, Joel and Rainforth, Tom and Zhang, Cheng and Foster, Adam}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {14445--14464}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/ivanova23a/ivanova23a.pdf}, url = {https://proceedings.mlr.press/v202/ivanova23a.html}, abstract = {We formalize the problem of contextual optimization through the lens of Bayesian experimental design and propose CO-BED—a general, model-agnostic framework for designing contextual experiments using information-theoretic principles. After formulating a suitable information-based objective, we employ black-box variational methods to simultaneously estimate it and optimize the designs in a single stochastic gradient scheme. In addition, to accommodate discrete actions within our framework, we propose leveraging continuous relaxation schemes, which can naturally be integrated into our variational objective. As a result, CO-BED provides a general and automated solution to a wide range of contextual optimization problems. We illustrate its effectiveness in a number of experiments, where CO-BED demonstrates competitive performance even when compared to bespoke, model-specific alternatives.} }
Endnote
%0 Conference Paper %T CO-BED: Information-Theoretic Contextual Optimization via Bayesian Experimental Design %A Desi R. Ivanova %A Joel Jennings %A Tom Rainforth %A Cheng Zhang %A Adam Foster %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-ivanova23a %I PMLR %P 14445--14464 %U https://proceedings.mlr.press/v202/ivanova23a.html %V 202 %X We formalize the problem of contextual optimization through the lens of Bayesian experimental design and propose CO-BED—a general, model-agnostic framework for designing contextual experiments using information-theoretic principles. After formulating a suitable information-based objective, we employ black-box variational methods to simultaneously estimate it and optimize the designs in a single stochastic gradient scheme. In addition, to accommodate discrete actions within our framework, we propose leveraging continuous relaxation schemes, which can naturally be integrated into our variational objective. As a result, CO-BED provides a general and automated solution to a wide range of contextual optimization problems. We illustrate its effectiveness in a number of experiments, where CO-BED demonstrates competitive performance even when compared to bespoke, model-specific alternatives.
APA
Ivanova, D.R., Jennings, J., Rainforth, T., Zhang, C. & Foster, A.. (2023). CO-BED: Information-Theoretic Contextual Optimization via Bayesian Experimental Design. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:14445-14464 Available from https://proceedings.mlr.press/v202/ivanova23a.html.

Related Material