Counterexample-guided inductive synthesis for probabilistic systems

被引:5
|
作者
Ceska, Milan [1 ]
Hensel, Christian [2 ]
Junges, Sebastian [3 ]
Katoen, Joost-Pieter [2 ]
机构
[1] Brno Univ Technol, Brno, Czech Republic
[2] Rhein Westfal TH Aachen, Aachen, Germany
[3] Univ Calif Berkeley, Berkeley, CA 94720 USA
关键词
Program Synthesis; Markov Chains; Probabilistic Model Checking; Counterexamples; CEGIS; QUANTITATIVE VERIFICATION; PARAMETER SYNTHESIS; MODELS;
D O I
10.1007/s00165-021-00547-2
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
This paper presents counterexample-guided inductive synthesis (CEGIS) to automatically synthesise probabilistic models. The starting point is a family of finite-stateMarkov chains with related but distinct topologies. Such families can succinctly be described by a sketch of a probabilistic program. Program sketches are programs containing holes. Every hole has a finite repertoire of possible program snippets by which it can be filled.We study several synthesis problems-feasibility, optimal synthesis, and complete partitioning-for a given quantitative specification phi. Feasibility amounts to determine a family member satisfying., optimal synthesis amounts to find a family member that maximises the probability to satisfy., and complete partitioning splits the family in satisfying and refuting members. Each of these problems can be considered under the additional constraint of minimising the total cost of instantiations, e.g., what are all possible instantiations for phi that are within a certain budget? The synthesis problems are tackled using a CEGIS approach. The crux is to aggressively prune the search space by using counterexamples provided by a probabilistic model checker. Counterexamples can be viewed as sub-Markov chains that rule out all family members that share this sub-chain. Our CEGIS approach leverages efficient probabilistic model checking, modernSMTsolving, and programsnippets as counterexamples. Experiments on case studies froma diverse nature-controller synthesis, program sketching, and security-show that synthesis among up to a million candidate designs can be done using a few thousand verification queries.
引用
收藏
页码:637 / 667
页数:31
相关论文
共 50 条
  • [31] Counterexample-Guided Correlation Algorithm for Translation Validation
    Gupta, Shubhani
    Rose, Abhishek
    Bansal, Sorav
    PROCEEDINGS OF THE ACM ON PROGRAMMING LANGUAGES-PACMPL, 2020, 4 (OOPSLA):
  • [32] Counterexample-guided computation of polyhedral Lyapunov functions for piecewise linear systems
    Berger, Guillaume O.
    Sankaranarayanan, Sriram
    AUTOMATICA, 2023, 155
  • [33] Effective Heuristics for Counterexample-Guided Abstraction Refinement
    He, Fei
    Song, Xiaoyu
    Gu, Ming
    Sun, Jiaguang
    GLSVLSI'07: PROCEEDINGS OF THE 2007 ACM GREAT LAKES SYMPOSIUM ON VLSI, 2007, : 393 - 398
  • [34] Automated Synthesis of Generalized Invariant Strategies via Counterexample-Guided Strategy Refinement
    Luo, Kailun
    Liu, Yongmei
    THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / THE TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2022, : 5800 - 5808
  • [35] Counterexample-guided abstraction refinement for symbolic model checking
    Clarke, E
    Grumberg, O
    Jha, S
    Lu, Y
    Veith, H
    JOURNAL OF THE ACM, 2003, 50 (05) : 752 - 794
  • [36] Counterexample-guided abstraction refinement for symmetric concurrent programs
    Donaldson, Alastair F.
    Kaiser, Alexander
    Kroening, Daniel
    Tautschnig, Michael
    Wahl, Thomas
    FORMAL METHODS IN SYSTEM DESIGN, 2012, 41 (01) : 25 - 44
  • [37] Counterexample-guided abstraction refinement for linear programs with arrays
    Alessandro Armando
    Massimo Benerecetti
    Jacopo Mantovani
    Automated Software Engineering, 2014, 21 : 225 - 285
  • [38] Solving quantified linear arithmetic by counterexample-guided instantiation
    Reynolds, Andrew
    King, Tim
    Kuncak, Viktor
    FORMAL METHODS IN SYSTEM DESIGN, 2017, 51 (03) : 500 - 532
  • [39] Counterexample-guided abstraction refinement for linear programs with arrays
    Armando, Alessandro
    Benerecetti, Massimo
    Mantovani, Jacopo
    AUTOMATED SOFTWARE ENGINEERING, 2014, 21 (02) : 225 - 285
  • [40] Solving quantified linear arithmetic by counterexample-guided instantiation
    Andrew Reynolds
    Tim King
    Viktor Kuncak
    Formal Methods in System Design, 2017, 51 : 500 - 532