AquaSense: Automated Sensitivity Analysis of Probabilistic Programs via Quantized Inference

Zitong Zhou, Zixin Huang, Sasa Misailovic

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We propose a novel tool, AquaSense, to automatically reason about the sensitivity analysis of probabilistic programs. In the context of probabilistic programs, sensitivity analysis investigates how the perturbation in the parameters of prior distributions affects the program’s result, i.e., the program’s posterior distribution. AquaSense leverages quantized inference, an efficient and accurate approximate inference algorithm that represents distributions of random variables with quantized intervals. AquaSense is the first tool to support sensitivity analysis of probabilistic programs that is at the same time symbolic, differentiable, and practical. Our evaluation compares AquaSense with an existing system PSense (a system that relies on fully symbolic inference). AquaSense can compute the sensitivity of all 45 parameters from 12 programs, compared to 11/45 that PSense computes. AquaSense is particularly effective on programs with continuous distributions: it achieves an average speedup of 18.10 × over PSense (which, in contrast, can solve only a handful of problems). Our evaluation shows that AquaSense computes exact results on discrete programs. On 91% of evaluated continuous parameters, AquaSense computed the sensitivity results within 40 s with high accuracy (below 5% error). The paper also discusses AquaSense’s performance-accuracy trade-offs, which can enable different operational points for programs with different input data sizes.

Original languageEnglish (US)
Title of host publicationAutomated Technology for Verification and Analysis - 21st International Symposium, ATVA 2023, Proceedings
EditorsÉtienne André, Jun Sun
PublisherSpringer
Pages288-301
Number of pages14
ISBN (Print)9783031453311
DOIs
StatePublished - 2023
Event21st International Symposium on Automated Technology for Verification and Analysis, ATVA 2023 - Singapore, Singapore
Duration: Oct 24 2023Oct 27 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14216 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference21st International Symposium on Automated Technology for Verification and Analysis, ATVA 2023
Country/TerritorySingapore
CitySingapore
Period10/24/2310/27/23

Keywords

  • Probabilistic Programming
  • Quantized Inference
  • Sensitivity Analysis

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'AquaSense: Automated Sensitivity Analysis of Probabilistic Programs via Quantized Inference'. Together they form a unique fingerprint.

Cite this