Projekteja vuodessa
Abstrakti
Approximate Bayesian computation (ABC) is a popular likelihood-free inference method for models with intractable likelihood functions. As ABC methods usually rely on comparing summary statistics of observed and simulated data, the choice of the statistics is crucial. This choice involves a trade-off between loss of information and dimensionality reduction, and is often determined based on domain knowledge. However, handcrafting and selecting suitable statistics is a laborious task involving multiple trial-and-error steps. In this work, we introduce an active learning method for ABC statistics selection which reduces the domain expert’s work considerably. By involving the experts, we are able to handle misspecified models, unlike the existing dimension reduction methods. Moreover, empirical results show better posterior estimates than with existing methods, when the simulation budget is limited.
Alkuperäiskieli | Englanti |
---|---|
Otsikko | Proceedings of the 39th International Conference on Machine Learning |
Kustantaja | JMLR |
Sivut | 1893-1905 |
Tila | Julkaistu - 2022 |
OKM-julkaisutyyppi | A4 Artikkeli konferenssijulkaisussa |
Tapahtuma | International Conference on Machine Learning - Baltimore, Yhdysvallat Kesto: 17 heinäk. 2022 → 23 heinäk. 2022 Konferenssinumero: 39 |
Julkaisusarja
Nimi | Proceedings of Machine Learning Research |
---|---|
Kustantaja | PMLR |
Vuosikerta | 162 |
ISSN (elektroninen) | 2640-3498 |
Conference
Conference | International Conference on Machine Learning |
---|---|
Lyhennettä | ICML |
Maa/Alue | Yhdysvallat |
Kaupunki | Baltimore |
Ajanjakso | 17/07/2022 → 23/07/2022 |
Sormenjälki
Sukella tutkimusaiheisiin 'Approximate Bayesian Computation with Domain Expert in the Loop'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.Projektit
- 1 Päättynyt
-
-: Finnish Center for Artificial Intelligence
Kaski, S. (Vastuullinen tutkija)
01/01/2019 → 31/12/2022
Projekti: Academy of Finland: Other research funding