Declarative probabilistic programming with Datalog
<p>Probabilistic programming languages are used for developing statistical models. They typically consist of two components: a specification of a stochastic process (the prior), and a specification of observations that restrict the probability space to a conditional subspace (the poste...
Main Authors: | , , , , |
---|---|
Format: | Journal article |
Published: |
Association for Computing Machinery
2017
|
_version_ | 1797095844838637568 |
---|---|
author | Barany, V Ten Cate, B Kimelfeld, b Olteanu, D Vagena, Z |
author_facet | Barany, V Ten Cate, B Kimelfeld, b Olteanu, D Vagena, Z |
author_sort | Barany, V |
collection | OXFORD |
description | <p>Probabilistic programming languages are used for developing statistical models. They typically consist of two components: a specification of a stochastic process (the prior), and a specification of observations that restrict the probability space to a conditional subspace (the posterior). Use cases of such formalisms include the development of algorithms in machine learning and artificial intelligence. </p> <p>In this article we establish a probabilistic-programming extension of Datalog that, on the one hand, allows for defining a rich family of statistical models, and on the other hand retains the fundamental properties of declarativity. Our proposed extension provides mechanisms to include common numerical probability functions; in particular, conclusions of rules may contain values drawn from such functions. The semantics of a program is a probability distribution over the possible outcomes of the input database with respect to the program. Observations are naturally incorporated by means of integrity constraints over the extensional and intensional relations. The resulting semantics is robust under different chases and invariant to rewritings that preserve logical equivalence.</p> |
first_indexed | 2024-03-07T04:33:46Z |
format | Journal article |
id | oxford-uuid:cf391c23-6078-44b6-b24d-66c6acdc1614 |
institution | University of Oxford |
last_indexed | 2024-03-07T04:33:46Z |
publishDate | 2017 |
publisher | Association for Computing Machinery |
record_format | dspace |
spelling | oxford-uuid:cf391c23-6078-44b6-b24d-66c6acdc16142022-03-27T07:41:01ZDeclarative probabilistic programming with DatalogJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:cf391c23-6078-44b6-b24d-66c6acdc1614Symplectic Elements at OxfordAssociation for Computing Machinery2017Barany, VTen Cate, BKimelfeld, bOlteanu, DVagena, Z<p>Probabilistic programming languages are used for developing statistical models. They typically consist of two components: a specification of a stochastic process (the prior), and a specification of observations that restrict the probability space to a conditional subspace (the posterior). Use cases of such formalisms include the development of algorithms in machine learning and artificial intelligence. </p> <p>In this article we establish a probabilistic-programming extension of Datalog that, on the one hand, allows for defining a rich family of statistical models, and on the other hand retains the fundamental properties of declarativity. Our proposed extension provides mechanisms to include common numerical probability functions; in particular, conclusions of rules may contain values drawn from such functions. The semantics of a program is a probability distribution over the possible outcomes of the input database with respect to the program. Observations are naturally incorporated by means of integrity constraints over the extensional and intensional relations. The resulting semantics is robust under different chases and invariant to rewritings that preserve logical equivalence.</p> |
spellingShingle | Barany, V Ten Cate, B Kimelfeld, b Olteanu, D Vagena, Z Declarative probabilistic programming with Datalog |
title | Declarative probabilistic programming with Datalog |
title_full | Declarative probabilistic programming with Datalog |
title_fullStr | Declarative probabilistic programming with Datalog |
title_full_unstemmed | Declarative probabilistic programming with Datalog |
title_short | Declarative probabilistic programming with Datalog |
title_sort | declarative probabilistic programming with datalog |
work_keys_str_mv | AT baranyv declarativeprobabilisticprogrammingwithdatalog AT tencateb declarativeprobabilisticprogrammingwithdatalog AT kimelfeldb declarativeprobabilisticprogrammingwithdatalog AT olteanud declarativeprobabilisticprogrammingwithdatalog AT vagenaz declarativeprobabilisticprogrammingwithdatalog |