Looping in the Human: Collaborative and Explainable Bayesian Optimization

Masaki Adachi, Brady Planden, David Howey, Michael A. Osborne, Sebastian Orbell, Natalia Ares, Krikamol Muandet, Siu Lun Chau
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:505-513, 2024.

Abstract

Like many optimizers, Bayesian optimization often falls short of gaining user trust due to opacity. While attempts have been made to develop human-centric optimizers, they typically assume user knowledge is well-specified and error-free, employing users mainly as supervisors of the optimization process. We relax these assumptions and propose a more balanced human-AI partnership with our Collaborative and Explainable Bayesian Optimization (CoExBO) framework. Instead of explicitly requiring a user to provide a knowledge model, CoExBO employs preference learning to seamlessly integrate human insights into the optimization, resulting in algorithmic suggestions that resonate with user preference. CoExBO explains its candidate selection every iteration to foster trust, empowering users with a clearer grasp of the optimization. Furthermore, CoExBO offers a no-harm guarantee, allowing users to make mistakes; even with extreme adversarial interventions, the algorithm converges asymptotically to a vanilla Bayesian optimization. We validate CoExBO’s efficacy through human-AI teaming experiments in lithium-ion battery design, highlighting substantial improvements over conventional methods. Code is available https://github.com/ma921/CoExBO.

Cite this Paper


BibTeX
@InProceedings{pmlr-v238-adachi24a, title = {Looping in the Human: Collaborative and Explainable {B}ayesian Optimization}, author = {Adachi, Masaki and Planden, Brady and Howey, David and A. Osborne, Michael and Orbell, Sebastian and Ares, Natalia and Muandet, Krikamol and Lun Chau, Siu}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {505--513}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/adachi24a/adachi24a.pdf}, url = {https://proceedings.mlr.press/v238/adachi24a.html}, abstract = {Like many optimizers, Bayesian optimization often falls short of gaining user trust due to opacity. While attempts have been made to develop human-centric optimizers, they typically assume user knowledge is well-specified and error-free, employing users mainly as supervisors of the optimization process. We relax these assumptions and propose a more balanced human-AI partnership with our Collaborative and Explainable Bayesian Optimization (CoExBO) framework. Instead of explicitly requiring a user to provide a knowledge model, CoExBO employs preference learning to seamlessly integrate human insights into the optimization, resulting in algorithmic suggestions that resonate with user preference. CoExBO explains its candidate selection every iteration to foster trust, empowering users with a clearer grasp of the optimization. Furthermore, CoExBO offers a no-harm guarantee, allowing users to make mistakes; even with extreme adversarial interventions, the algorithm converges asymptotically to a vanilla Bayesian optimization. We validate CoExBO’s efficacy through human-AI teaming experiments in lithium-ion battery design, highlighting substantial improvements over conventional methods. Code is available https://github.com/ma921/CoExBO.} }
Endnote
%0 Conference Paper %T Looping in the Human: Collaborative and Explainable Bayesian Optimization %A Masaki Adachi %A Brady Planden %A David Howey %A Michael A. Osborne %A Sebastian Orbell %A Natalia Ares %A Krikamol Muandet %A Siu Lun Chau %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-adachi24a %I PMLR %P 505--513 %U https://proceedings.mlr.press/v238/adachi24a.html %V 238 %X Like many optimizers, Bayesian optimization often falls short of gaining user trust due to opacity. While attempts have been made to develop human-centric optimizers, they typically assume user knowledge is well-specified and error-free, employing users mainly as supervisors of the optimization process. We relax these assumptions and propose a more balanced human-AI partnership with our Collaborative and Explainable Bayesian Optimization (CoExBO) framework. Instead of explicitly requiring a user to provide a knowledge model, CoExBO employs preference learning to seamlessly integrate human insights into the optimization, resulting in algorithmic suggestions that resonate with user preference. CoExBO explains its candidate selection every iteration to foster trust, empowering users with a clearer grasp of the optimization. Furthermore, CoExBO offers a no-harm guarantee, allowing users to make mistakes; even with extreme adversarial interventions, the algorithm converges asymptotically to a vanilla Bayesian optimization. We validate CoExBO’s efficacy through human-AI teaming experiments in lithium-ion battery design, highlighting substantial improvements over conventional methods. Code is available https://github.com/ma921/CoExBO.
APA
Adachi, M., Planden, B., Howey, D., A. Osborne, M., Orbell, S., Ares, N., Muandet, K. & Lun Chau, S.. (2024). Looping in the Human: Collaborative and Explainable Bayesian Optimization. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:505-513 Available from https://proceedings.mlr.press/v238/adachi24a.html.

Related Material