Modular Probabilistic Models via Algebraic Effects

Minh H Nguyen; Roly N T Perera; Meng Wang; Nicolas Wu

doi:10.1145/3547635

Modular Probabilistic Models via Algebraic Effects

Minh H Nguyen, Roly N T Perera, Meng Wang, Nicolas Wu

Research output: Contribution to journal › Article (Academic Journal) › peer-review

3 Citations (Scopus)

36 Downloads (Pure)

Abstract

Probabilistic programming languages (PPLs) allow programmers to construct statistical models and then simulate data or perform inference over them. Many PPLs restrict models to a particular instance of simulation or inference, limiting their reusability. In other PPLs, models are not readily composable. Using Haskell as the host language, we present an embedded domain specific language based on algebraic effects, where probabilistic models are modular, first-class, and reusable for both simulation and inference. We also demonstrate how simulation and inference can be expressed naturally as composable program transformations using algebraic effect handlers.

Original language	English
Article number	104
Pages (from-to)	381-410
Number of pages	30
Journal	Proceedings of the ACM on Programming Languages
Volume	6
Issue number	ICFP
Early online date	31 Aug 2022
DOIs	https://doi.org/10.1145/3547635
Publication status	Published - 11 Sept 2022

Bibliographical note

Funding Information:
Our sincere thanks to Matija Pretnar for shepherding this paper during the conditional acceptance phase, and to the anonymous reviewers for their valuable discussion. We also thank Alessio Zakaria who has been a constant source of support and drive behind our ideas, and the members of the Bristol Programming Languages research group for creating the wonderful environment for this work. This work is partly supported by the EPSRC grant EXHIBIT : Expressive High-Level Languages for Bidirectional Transformations (EP/T008911/1).

Publisher Copyright:
© 2022 Copyright held by the owner/author(s).

Research Groups and Themes

Programming Languages

Access to Document

10.1145/3547635Licence: CC BY

Full-text PDF (final published version)
This is the final published version of the article (version of record). It first appeared online via ACM at https://doi.org/10.5281/zenodo.6651953. Please refer to any applicable terms of use of the publisher.
Final published version, 553 KBLicence: CC BY

Handle.net

Persistent link

Effects and effect handlers for probabilistic programming
Nguyen, M. H. (Author), Wang, M. (Supervisor) & Perera, R. (Supervisor), 5 Dec 2023
Student thesis: Doctoral Thesis › Doctor of Philosophy (PhD)

File

Cite this

@article{216dda928d2b4e6eb9daad10967c8a80,

title = "Modular Probabilistic Models via Algebraic Effects",

abstract = "Probabilistic programming languages (PPLs) allow programmers to construct statistical models and then simulate data or perform inference over them. Many PPLs restrict models to a particular instance of simulation or inference, limiting their reusability. In other PPLs, models are not readily composable. Using Haskell as the host language, we present an embedded domain specific language based on algebraic effects, where probabilistic models are modular, first-class, and reusable for both simulation and inference. We also demonstrate how simulation and inference can be expressed naturally as composable program transformations using algebraic effect handlers.",

author = "Nguyen, {Minh H} and Perera, {Roly N T} and Meng Wang and Nicolas Wu",

note = "Funding Information: Our sincere thanks to Matija Pretnar for shepherding this paper during the conditional acceptance phase, and to the anonymous reviewers for their valuable discussion. We also thank Alessio Zakaria who has been a constant source of support and drive behind our ideas, and the members of the Bristol Programming Languages research group for creating the wonderful environment for this work. This work is partly supported by the EPSRC grant EXHIBIT : Expressive High-Level Languages for Bidirectional Transformations (EP/T008911/1). Publisher Copyright: {\textcopyright} 2022 Copyright held by the owner/author(s).",

year = "2022",

month = sep,

day = "11",

doi = "10.1145/3547635",

language = "English",

volume = "6",

pages = "381--410",

journal = "Proceedings of the ACM on Programming Languages",

issn = "2475-1421",

publisher = "Association for Computing Machinery (ACM)",

number = "ICFP",

}

TY - JOUR

T1 - Modular Probabilistic Models via Algebraic Effects

AU - Nguyen, Minh H

AU - Perera, Roly N T

AU - Wang, Meng

AU - Wu, Nicolas

N1 - Funding Information: Our sincere thanks to Matija Pretnar for shepherding this paper during the conditional acceptance phase, and to the anonymous reviewers for their valuable discussion. We also thank Alessio Zakaria who has been a constant source of support and drive behind our ideas, and the members of the Bristol Programming Languages research group for creating the wonderful environment for this work. This work is partly supported by the EPSRC grant EXHIBIT : Expressive High-Level Languages for Bidirectional Transformations (EP/T008911/1). Publisher Copyright: © 2022 Copyright held by the owner/author(s).

PY - 2022/9/11

Y1 - 2022/9/11

N2 - Probabilistic programming languages (PPLs) allow programmers to construct statistical models and then simulate data or perform inference over them. Many PPLs restrict models to a particular instance of simulation or inference, limiting their reusability. In other PPLs, models are not readily composable. Using Haskell as the host language, we present an embedded domain specific language based on algebraic effects, where probabilistic models are modular, first-class, and reusable for both simulation and inference. We also demonstrate how simulation and inference can be expressed naturally as composable program transformations using algebraic effect handlers.

AB - Probabilistic programming languages (PPLs) allow programmers to construct statistical models and then simulate data or perform inference over them. Many PPLs restrict models to a particular instance of simulation or inference, limiting their reusability. In other PPLs, models are not readily composable. Using Haskell as the host language, we present an embedded domain specific language based on algebraic effects, where probabilistic models are modular, first-class, and reusable for both simulation and inference. We also demonstrate how simulation and inference can be expressed naturally as composable program transformations using algebraic effect handlers.

U2 - 10.1145/3547635

DO - 10.1145/3547635

M3 - Article (Academic Journal)

SN - 2475-1421

VL - 6

SP - 381

EP - 410

JO - Proceedings of the ACM on Programming Languages

JF - Proceedings of the ACM on Programming Languages

IS - ICFP

M1 - 104

ER -

Modular Probabilistic Models via Algebraic Effects

Abstract

Bibliographical note

Research Groups and Themes

Access to Document

Handle.net

Fingerprint

Student theses

Effects and effect handlers for probabilistic programming

Cite this