elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Verification of Sigmoidal Artificial Neural Networks using iSAT

Grundt, Dominik and Jurj, Sorin Liviu and Hagemann, Willem and Kröger, Paul and Fränzle, Martin (2022) Verification of Sigmoidal Artificial Neural Networks using iSAT. In: 7th International Workshop on Symbolic-Numeric Methods for Reasoning about CPS and IoT, SNR 2021, 361, pp. 45-60. Electronic Proceedings in Theoretical Computer Science. 7th International Workshop on Symbolic-Numeric Methods for Reasoning about CPS and IoT, 2021-08-23, Online. doi: 10.4204/EPTCS.361.6. ISSN 2075-2180.

[img] PDF
477kB

Official URL: https://dx.doi.org/10.4204/EPTCS.361.6

Abstract

This paper presents an approach for verifying the behaviour of nonlinear Artificial Neural Networks (ANNs) found in cyber-physical safety-critical systems. We implement a dedicated interval constraint propagator for the sigmoid function into the SMT solver iSAT and compare this approach with a compositional approach encoding the sigmoid function by basic arithmetic features available in iSAT and an approximating approach. Our experimental results show that the dedicated and the compositional approach clearly outperform the approximating approach. Throughout all our benchmarks, the dedicated approach showed an equal or better performance compared to the compositional approach.

Item URL in elib:https://elib.dlr.de/187452/
Document Type:Conference or Workshop Item (Speech)
Additional Information:Das Paper wurde im Rahmen des Projektes KI Wissen (gefördert durch BMWK - grant agreement No. 19A20020M) und des Projektes ViVre (State of Lower Saxony within the framework “Zukunftslabor Mobilit) erarbeitet. SCOPUS Eintrag/Ranking sollte zeitnah erfolgen.
Title:Verification of Sigmoidal Artificial Neural Networks using iSAT
Authors:
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Grundt, DominikUNSPECIFIEDhttps://orcid.org/0000-0002-8233-7429UNSPECIFIED
Jurj, Sorin LiviuUNSPECIFIEDhttps://orcid.org/0000-0002-2689-6015UNSPECIFIED
Hagemann, WillemUNSPECIFIEDhttps://orcid.org/0000-0002-9138-8340UNSPECIFIED
Kröger, PaulUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Fränzle, MartinUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Date:13 July 2022
Journal or Publication Title:7th International Workshop on Symbolic-Numeric Methods for Reasoning about CPS and IoT, SNR 2021
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:No
Volume:361
DOI:10.4204/EPTCS.361.6
Page Range:pp. 45-60
Editors:
EditorsEmailEditor's ORCID iDORCID Put Code
Remke, AnneUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Tran, Dung HoangUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Publisher:Electronic Proceedings in Theoretical Computer Science
ISSN:2075-2180
Status:Published
Keywords:AI Verification, nonlinear activation function
Event Title:7th International Workshop on Symbolic-Numeric Methods for Reasoning about CPS and IoT
Event Location:Online
Event Type:Workshop
Event Date:23 August 2021
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Transport
HGF - Program Themes:Road Transport
DLR - Research area:Transport
DLR - Program:V ST Straßenverkehr
DLR - Research theme (Project):V - V&V4NGC - Methoden, Prozesse und Werkzeugketten für die Validierung & Verifikation von NGC
Location: Oldenburg
Institutes and Institutions:Institute of Systems Engineering for Future Mobility > Systems Theory and Design
Deposited By: Grundt, Dominik
Deposited On:09 Aug 2022 08:49
Last Modified:24 Apr 2024 20:48

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.