Publication Details

Inductive Synthesis for Probabilistic Programs Reaches New Horizons

ANDRIUSHCHENKO Roman, ČEŠKA Milan, JUNGES Sebastian and KATOEN Joost-Pieter. Inductive Synthesis for Probabilistic Programs Reaches New Horizons. In: International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS). Lecture Notes in Computer Science. Cham: Springer International Publishing, 2021, pp. 191-209. ISBN 978-3-030-72015-5.
Czech title
Nové pokroky v syntéze pravděpodobnostních programů
Type
conference paper
Language
english
Authors
Andriushchenko Roman, Ing. (DITS FIT BUT)
Češka Milan, doc. RNDr., Ph.D. (DITS FIT BUT)
Junges Sebastian (RWTH Aachen University)
Katoen Joost-Pieter (RWTH)
Keywords

Probabilistic programs,
Inductive Synthesis,
Counterexamples,
Probabilistic Model Checking

Abstract

This paper presents a novel method for the automated synthesis of probabilistic programs. The starting point is a program sketch representing a finite family of finite-state Markov chains with related but distinct topologies, and a reachability specification. The method builds on a novel inductive oracle that greedily generates counter-examples (CEs) for violating programs and uses them to prune the family. These CEs leverage the semantics of the family in the form of bounds on its best- and worst-case behaviour provided by a deductive oracle using an MDP abstraction. The method further monitors the performance of the synthesis and adaptively switches between inductive and deductive reasoning. Our experiments demonstrate that the novel CE construction provides a significantly faster and more effective pruning strategy leading to an accelerated synthesis process on a wide range of benchmarks. For challenging problems, such as the synthesis of decentralized partially-observable controllers, we reduce the run-time from a day to minutes.

Published
2021
Pages
191-209
Proceedings
International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS)
Series
Lecture Notes in Computer Science
Conference
European Joint Conferences on Theory and Practice of Software -- ETAPS'21 (TACAS'21), Lucemburk, LU
ISBN
978-3-030-72015-5
Publisher
Springer International Publishing
Place
Cham, DE
DOI
BibTeX
@INPROCEEDINGS{FITPUB12498,
   author = "Roman Andriushchenko and Milan \v{C}e\v{s}ka and Sebastian Junges and Joost-Pieter Katoen",
   title = "Inductive Synthesis for Probabilistic Programs Reaches New Horizons",
   pages = "191--209",
   booktitle = "International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS)",
   series = "Lecture Notes in Computer Science",
   year = 2021,
   location = "Cham, DE",
   publisher = "Springer International Publishing",
   ISBN = "978-3-030-72015-5",
   doi = "10.1007/978-3-030-72016-2\_11",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/12498"
}
Back to top