Publication Details
PAC Learning-Based Verification and Model Synthesis
Hsieh Chiao (FIT)
Lengál Ondřej, Ing., Ph.D. (DITS)
Lii Tsung-Ju (FIT)
Tsai Ming-Hsien (FIT)
Wang Bow-Yaw (FIT)
Wang Farn (FIT)
model synthesis, PAC learning, finite automata, program verification
We introduce a novel technique for verification and model synthesis of sequential programs. Our technique is based on learning an approximate regular model of the set of feasible paths in a program, and testing whether this model contains an incorrect behavior. Exact learning algorithms require checking equivalence between the model and the program, which is a difficult problem, in general undecidable. Our learning procedure is therefore based on the framework of probably approximately correct (PAC) learning, which uses sampling instead, and provides correctness guarantees expressed using the terms error probability and confidence. Besides the verification result, our procedure also outputs the model with the said correctness guarantees. Obtained preliminary experiments show encouraging results, in some cases even outperforming mature software verifiers.
@inproceedings{BUT130941,
author="Yu-Fang {Chen} and Chiao {Hsieh} and Ondřej {Lengál} and Tsung-Ju {Lii} and Ming-Hsien {Tsai} and Bow-Yaw {Wang} and Farn {Wang}",
title="PAC Learning-Based Verification and Model Synthesis",
booktitle="Proceedings of the 38th International Conference on Software Engineering",
year="2016",
pages="714--724",
publisher="Association for Computing Machinery",
address="Austin, TX",
doi="10.1145/2884781.2884860",
isbn="978-1-4503-3900-1",
url="http://dx.doi.org/10.1145/2884781.2884860"
}