Automatic Abstraction Refinement in Neural Network Verification using Sensitivity Analysis

DOI10.1145/3575870.3587129MaRDI QIDQ6202102zbMATH OpenOpenAlexWikidataFDO

Authors Matthias Althoff

Publication date 21 February 2024

Published in Proceedings of the 26th ACM International Conference on Hybrid Systems: Computation and Control (Search for Journal in Brave)

Full work available at URL https://mediatum.ub.tum.de/doc/1710698/document.pdf

zbMATH Keywords

automatic abstraction refinementformal verificationneural networkspolynomial zonotopes.sensitivity analysisset-based computing

Mathematics Subject Classification ID

Formal languages and automata (68Q45) Specification and verification (program logics, model checking, etc.) (68Q60) Control/observation systems governed by functional relations other than differential equations (such as hybrid and switching systems) (93C30)

Cites work

• An empirical study of QBF encodings: from treewidth estimation to useful preprocessing
• Branch and bound for piecewise linear neural network verification
• scientific article; zbMATH DE number 1670775 (Why is no real title available?)
• Hybrid Systems: Computation and Control
• JuliaReach
• Learning representations by back-propagating errors
• Multilayer feedforward networks are universal approximators
• Pattern recognition and machine learning.
• Reachable sets of classifiers and regression models: (non-)robustness analysis and robust training
• Reluplex: an efficient SMT solver for verifying deep neural networks
• Safety verification of deep neural networks
• Sparse Polynomial Zonotopes: A Novel Set Representation for Reachability Analysis
• Suppression of the wrapping effect by Taylor model-based verified integrators: long-term stabilization by preconditioning
• Utilizing dependencies to obtain subsets of reachable sets
• Verification of piecewise deep neural networks: a star set approach with zonotope pre-filter
• Verisig 2.0: verification of neural network controllers using Taylor model preconditioning

Cited in

(1)

• Distributionally robust statistical verification with imprecise neural networks

This page was built for publication: Automatic Abstraction Refinement in Neural Network Verification using Sensitivity Analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6202102)