English

Scenario-Based Verification of Uncertain MDPs

Logic in Computer Science 2020-02-26 v2 Optimization and Control

Abstract

We consider Markov decision processes (MDPs) in which the transition probabilities and rewards belong to an uncertainty set parametrized by a collection of random variables. The probability distributions for these random parameters are unknown. The problem is to compute the probability to satisfy a temporal logic specification within any MDP that corresponds to a sample from these unknown distributions. In general, this problem is undecidable, and we resort to techniques from so-called scenario optimization. Based on a finite number of samples of the uncertain parameters, each of which induces an MDP, the proposed method estimates the probability of satisfying the specification by solving a finite-dimensional convex optimization problem. The number of samples required to obtain a high confidence on this estimate is independent from the number of states and the number of random parameters. Experiments on a large set of benchmarks show that a few thousand samples suffice to obtain high-quality confidence bounds with a high probability.

Keywords

Cite

@article{arxiv.1912.11223,
  title  = {Scenario-Based Verification of Uncertain MDPs},
  author = {Murat Cubuktepe and Nils Jansen and Sebastian Junges and Joost-Pieter Katoen and Ufuk Topcu},
  journal= {arXiv preprint arXiv:1912.11223},
  year   = {2020}
}

Comments

Accepted to TACAS 2020

R2 v1 2026-06-23T12:55:25.928Z