Central Moment Analysis for Cost Accumulators in Probabilistic Programs

被引:14
作者
Wang, Di [1 ]
Hoffmann, Jan [1 ]
Reps, Thomas [2 ]
机构
[1] Carnegie Mellon Univ, Pittsburgh, PA 15213 USA
[2] Univ Wisconsin, Madison, WI 53706 USA
来源
PROCEEDINGS OF THE 42ND ACM SIGPLAN INTERNATIONAL CONFERENCE ON PROGRAMMING LANGUAGE DESIGN AND IMPLEMENTATION (PLDI '21) | 2021年
基金
美国国家科学基金会;
关键词
Probabilistic programs; central moments; cost analysis; tail bounds;
D O I
10.1145/3453483.3454062
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
For probabilistic programs, it is usually not possible to automatically derive exact information about their properties, such as the distribution of states at a given program point. Instead, one can attempt to derive approximations, such as upper bounds on tail probabilities. Such bounds can be obtained via concentration inequalities, which rely on the moments of a distribution, such as the expectation (the first raw moment) or the variance (the second central moment). Tail bounds obtained using central moments are often tighter than the ones obtained using raw moments, but automatically analyzing central moments is more challenging. This paper presents an analysis for probabilistic programs that automatically derives symbolic upper and lower bounds on variances, as well as higher central moments, of cost accumulators. To overcome the challenges of higher-moment analysis, it generalizes analyses for expectations with an algebraic abstraction that simultaneously analyzes different moments, utilizing relations between them. A key innovation is the notion of moment-polymorphic recursion, and a practical derivation system that handles recursive functions. The analysis has been implemented using a template-based technique that reduces the inference of polynomial bounds to linear programming. Experiments with our prototype central-moment analyzer show that, despite the analyzer's upper/lower bounds on various quantities, it obtains tighter tail bounds than an existing system that uses only raw moments, such as expectations.
引用
收藏
页码:559 / 573
页数:15
相关论文
共 41 条
[1]  
Barthe G., 2018, EUROPEAN S PROGRAMMI, V10801, P117, DOI DOI 10.1007/978-3-319-89884-15
[2]   Synthesizing Probabilistic Invariants via Doob's Decomposition [J].
Barthe, Gilles ;
Espitau, Thomas ;
Ferrer Fioriti, Luis Maria ;
Hsu, Justin .
COMPUTER AIDED VERIFICATION, (CAV 2016), PT I, 2016, 9779 :43-61
[3]   Formal Certification of Code-Based Cryptographic Proofs [J].
Barthe, Gilles ;
Gregoire, Benjamin ;
Beguelin, Santiago Zanella .
ACM SIGPLAN NOTICES, 2009, 44 (01) :90-101
[4]   Automatic Generation of Moment-Based Invariants for Prob-Solvable Loops [J].
Bartocci, Ezio ;
Kovacs, Laura ;
Stankovic, Miroslav .
AUTOMATED TECHNOLOGY FOR VERIFICATION AND ANALYSIS (ATVA 2019), 2019, 11781 :255-276
[5]  
Batz Kevin, 2018, Programming Languages and Systems. 27th European Symposium on Programming, ESOP 2018, held as part of the European Joint Conferences on Theory and Practice of Software, ETAPS 2018. Proceedings: LNCS 10801, P186, DOI 10.1007/978-3-319-89884-1_7
[6]  
Borgström J, 2016, ACM SIGPLAN NOTICES, V51, P33, DOI [10.1145/3022670.2951942, 10.1145/2951913.2951942]
[7]   Uncertainty Propagation Using Probabilistic Affine Forms and Concentration of Measure Inequalities [J].
Bouissou, Olivier ;
Goubault, Eric ;
Putot, Sylvie ;
Chakarov, Aleksandar ;
Sankaranarayanan, Sriram .
TOOLS AND ALGORITHMS FOR THE CONSTRUCTION AND ANALYSIS OF SYSTEMS (TACAS 2016), 2016, 9636 :225-243
[8]   Automated Resource Analysis with Coq Proof Objects [J].
Carbonneaux, Quentin ;
Hoffmann, Jan ;
Reps, Thomas ;
Shao, Zhong .
COMPUTER AIDED VERIFICATION (CAV 2017), PT II, 2017, 10427 :64-85
[9]  
Chakarov Aleksandar, 2013, Computer Aided Verification. 25th International Conference, CAV 2013. Proceedings. LNCS 8044, P511, DOI 10.1007/978-3-642-39799-8_34
[10]   Termination Analysis of Probabilistic Programs Through Positivstellensatz's [J].
Chatterjee, Krishnendu ;
Fu, Hongfei ;
Goharshady, Amir Kafshdar .
COMPUTER AIDED VERIFICATION, (CAV 2016), PT I, 2016, 9779 :3-22