Nonlinear Bayesian optimal experimental design using logarithmic Sobolev inequalities
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911782758514688 |
|---|---|
| author | Li, Fengyi Belhadji, Ayoub Marzouk, Youssef |
| author_facet | Li, Fengyi Belhadji, Ayoub Marzouk, Youssef |
| contents | We study the problem of selecting $k$ experiments from a larger candidate pool, where the goal is to maximize mutual information (MI) between the selected subset and the underlying parameters. Finding the exact solution is to this combinatorial optimization problem is computationally costly, not only due to the complexity of the combinatorial search but also the difficulty of evaluating MI in nonlinear/non-Gaussian settings. We propose greedy approaches based on new computationally inexpensive lower bounds for MI, constructed via log-Sobolev inequalities. We demonstrate that our method outperforms random selection strategies, Gaussian approximations, and nested Monte Carlo (NMC) estimators of MI in various settings, including optimal design for nonlinear models with non-additive noise. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_15053 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Nonlinear Bayesian optimal experimental design using logarithmic Sobolev inequalities Li, Fengyi Belhadji, Ayoub Marzouk, Youssef Machine Learning Methodology We study the problem of selecting $k$ experiments from a larger candidate pool, where the goal is to maximize mutual information (MI) between the selected subset and the underlying parameters. Finding the exact solution is to this combinatorial optimization problem is computationally costly, not only due to the complexity of the combinatorial search but also the difficulty of evaluating MI in nonlinear/non-Gaussian settings. We propose greedy approaches based on new computationally inexpensive lower bounds for MI, constructed via log-Sobolev inequalities. We demonstrate that our method outperforms random selection strategies, Gaussian approximations, and nested Monte Carlo (NMC) estimators of MI in various settings, including optimal design for nonlinear models with non-additive noise. |
| title | Nonlinear Bayesian optimal experimental design using logarithmic Sobolev inequalities |
| topic | Machine Learning Methodology |
| url | https://arxiv.org/abs/2402.15053 |