Bayesian inference using data flow analysis

Guillaume Claret; Sriram K. Rajamani; Aditya V. Nori; Andrew D. Gordon; Johannes Borgström

doi:10.1145/2491411.2491423

Bayesian inference using data flow analysis

Guillaume Claret, Sriram K. Rajamani, Aditya V. Nori, Andrew D. Gordon, Johannes Borgström

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We present a new algorithm for Bayesian inference over probabilistic programs, based on data flow analysis techniques from the program analysis community. Unlike existing techniques for Bayesian inference on probabilistic programs, our data flow analysis algorithm is able to perform inference directly on probabilistic programs with loops. Even for loop-free programs, we show that data flow analysis offers better precision and better performance benefits over existing techniques. We also describe heuristics that are crucial for our inference to scale, and present an empirical evaluation of our algorithm over a range of benchmarks.

Original language	English
Title of host publication	Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013
Pages	92-102
Number of pages	11
DOIs	https://doi.org/10.1145/2491411.2491423
Publication status	Published - 2013

Access to Document

10.1145/2491411.2491423

http://doi.acm.org/10.1145/2491411.2491423

Cite this

Claret, G., Rajamani, S. K., Nori, A. V., Gordon, A. D., & Borgström, J. (2013). Bayesian inference using data flow analysis. In Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013 (pp. 92-102) https://doi.org/10.1145/2491411.2491423

@inproceedings{a21edc827d7f4879a96a34ff8c0c7f41,

title = "Bayesian inference using data flow analysis",

abstract = "We present a new algorithm for Bayesian inference over probabilistic programs, based on data flow analysis techniques from the program analysis community. Unlike existing techniques for Bayesian inference on probabilistic programs, our data flow analysis algorithm is able to perform inference directly on probabilistic programs with loops. Even for loop-free programs, we show that data flow analysis offers better precision and better performance benefits over existing techniques. We also describe heuristics that are crucial for our inference to scale, and present an empirical evaluation of our algorithm over a range of benchmarks. ",

author = "Guillaume Claret and Rajamani, {Sriram K.} and Nori, {Aditya V.} and Gordon, {Andrew D.} and Johannes Borgstr{\"o}m",

year = "2013",

doi = "10.1145/2491411.2491423",

language = "English",

isbn = "978-1-4503-2237-9",

pages = "92--102",

booktitle = "Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013",

}

Claret, G, Rajamani, SK, Nori, AV, Gordon, AD & Borgström, J 2013, Bayesian inference using data flow analysis. in Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013. pp. 92-102. https://doi.org/10.1145/2491411.2491423

Bayesian inference using data flow analysis. / Claret, Guillaume; Rajamani, Sriram K.; Nori, Aditya V. et al.
Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013. 2013. p. 92-102.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Bayesian inference using data flow analysis

AU - Claret, Guillaume

AU - Rajamani, Sriram K.

AU - Nori, Aditya V.

AU - Gordon, Andrew D.

AU - Borgström, Johannes

PY - 2013

Y1 - 2013

N2 - We present a new algorithm for Bayesian inference over probabilistic programs, based on data flow analysis techniques from the program analysis community. Unlike existing techniques for Bayesian inference on probabilistic programs, our data flow analysis algorithm is able to perform inference directly on probabilistic programs with loops. Even for loop-free programs, we show that data flow analysis offers better precision and better performance benefits over existing techniques. We also describe heuristics that are crucial for our inference to scale, and present an empirical evaluation of our algorithm over a range of benchmarks.

AB - We present a new algorithm for Bayesian inference over probabilistic programs, based on data flow analysis techniques from the program analysis community. Unlike existing techniques for Bayesian inference on probabilistic programs, our data flow analysis algorithm is able to perform inference directly on probabilistic programs with loops. Even for loop-free programs, we show that data flow analysis offers better precision and better performance benefits over existing techniques. We also describe heuristics that are crucial for our inference to scale, and present an empirical evaluation of our algorithm over a range of benchmarks.

U2 - 10.1145/2491411.2491423

DO - 10.1145/2491411.2491423

M3 - Conference contribution

SN - 978-1-4503-2237-9

SP - 92

EP - 102

BT - Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, ESEC/FSE'13, Saint Petersburg, Russian Federation, August 18-26, 2013

ER -

Bayesian inference using data flow analysis

Abstract

Access to Document

Fingerprint

Cite this