Looping in the Human Collaborative and Explainable Bayesian Optimization
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912070638764032 |
|---|---|
| author | Adachi, Masaki Planden, Brady Howey, David A. Osborne, Michael A. Orbell, Sebastian Ares, Natalia Muandet, Krikamol Chau, Siu Lun |
| author_facet | Adachi, Masaki Planden, Brady Howey, David A. Osborne, Michael A. Orbell, Sebastian Ares, Natalia Muandet, Krikamol Chau, Siu Lun |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2310_17273 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Looping in the Human Collaborative and Explainable Bayesian Optimization Adachi, Masaki Planden, Brady Howey, David A. Osborne, Michael A. Orbell, Sebastian Ares, Natalia Muandet, Krikamol Chau, Siu Lun Machine Learning Human-Computer Interaction 62C10, 62F15 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. |
| title | Looping in the Human Collaborative and Explainable Bayesian Optimization |
| topic | Machine Learning Human-Computer Interaction 62C10, 62F15 |
| url | https://arxiv.org/abs/2310.17273 |