Maximum a posteriori estimation by search in probabilistic programs
We introduce an approximate search algorithm for fast maximum a posteriori probability estimation in probabilistic programs, which we call Bayesian ascent Monte Carlo (BaMC). Probabilistic programs represent probabilistic models with varying number of mutually dependent finite, countable, and contin...
| Auteurs principaux: | , |
|---|---|
| Format: | Conference item |
| Publié: |
AAAI Publications
2015
|
| _version_ | 1826273017107316736 |
|---|---|
| author | Tolpin, D Wood, F |
| author_facet | Tolpin, D Wood, F |
| author_sort | Tolpin, D |
| collection | OXFORD |
| description | We introduce an approximate search algorithm for fast maximum a posteriori probability estimation in probabilistic programs, which we call Bayesian ascent Monte Carlo (BaMC). Probabilistic programs represent probabilistic models with varying number of mutually dependent finite, countable, and continuous random variables. BaMC is an anytime MAP search algorithm applicable to any combination of random variables and dependencies. We compare BaMC to other MAP estimation algorithms and show that BaMC is faster and more robust on a range of probabilistic models. |
| first_indexed | 2024-03-06T22:21:44Z |
| format | Conference item |
| id | oxford-uuid:554a0b51-9184-4396-9a48-819eee390c21 |
| institution | University of Oxford |
| last_indexed | 2024-03-06T22:21:44Z |
| publishDate | 2015 |
| publisher | AAAI Publications |
| record_format | dspace |
| spelling | oxford-uuid:554a0b51-9184-4396-9a48-819eee390c212022-03-26T16:43:04ZMaximum a posteriori estimation by search in probabilistic programsConference itemhttp://purl.org/coar/resource_type/c_5794uuid:554a0b51-9184-4396-9a48-819eee390c21Symplectic Elements at OxfordAAAI Publications2015Tolpin, DWood, FWe introduce an approximate search algorithm for fast maximum a posteriori probability estimation in probabilistic programs, which we call Bayesian ascent Monte Carlo (BaMC). Probabilistic programs represent probabilistic models with varying number of mutually dependent finite, countable, and continuous random variables. BaMC is an anytime MAP search algorithm applicable to any combination of random variables and dependencies. We compare BaMC to other MAP estimation algorithms and show that BaMC is faster and more robust on a range of probabilistic models. |
| spellingShingle | Tolpin, D Wood, F Maximum a posteriori estimation by search in probabilistic programs |
| title | Maximum a posteriori estimation by search in probabilistic programs |
| title_full | Maximum a posteriori estimation by search in probabilistic programs |
| title_fullStr | Maximum a posteriori estimation by search in probabilistic programs |
| title_full_unstemmed | Maximum a posteriori estimation by search in probabilistic programs |
| title_short | Maximum a posteriori estimation by search in probabilistic programs |
| title_sort | maximum a posteriori estimation by search in probabilistic programs |
| work_keys_str_mv | AT tolpind maximumaposterioriestimationbysearchinprobabilisticprograms AT woodf maximumaposterioriestimationbysearchinprobabilisticprograms |