Options
Concept Enforcement and Modularization as Methods for the ISO 26262 Safety Argumentation of Neural Networks
Schwalbe, Gesina; Schels, Martin (2020): „Concept Enforcement and Modularization as Methods for the ISO 26262 Safety Argumentation of Neural Networks“. In: Toulouse S. 1–10.
Faculty/Professorship:
Author:
Title of the compilation:
Proceeding of the 10th European Congress on Embedded Real Time Software and Systems
Corporate Body:
European Congress on Embedded Real Time Software and Systems (ERTS), 10, 2020, Toulouse
Publisher Information:
Year of publication:
2020
Pages:
Language:
English
Abstract:
Neural networks (NN) are prone to systematic faults which are hard to detect using the methods recommended by the ISO 26262 automotive functional safety standard. In this paper we propose a unified approach to two methods for NN safety argumentation: Assignment of human interpretable concepts to the internal representation of NNs to enable modularization and formal verification. Feasibility of the required concept embedding analysis is demonstrated in a minimal example and important aspects for generalization are investigated. The contribution of the methods is derived from a proposed generic argumentation structure for a NN model safety case.
GND Keywords: ;  ;  ;  ; 
ISO/DIS 26262
Funktionssicherheit
Maschinelles Lernen
Netzwerk
Künstliche Intelligenz
Keywords: ;  ;  ;  ;  ;  ; 
concept enforcement
machine learning
neural networks
functional safety
ISO 26262
goal structuring notation
explainable AI
DDC Classification:
RVK Classification:
Peer Reviewed:
Yes:
International Distribution:
Yes:
Open Access Journal:
Yes:
Type:
Conferenceobject
published:
June 2, 2020
Versioning
Question on publication
Permalink
https://fis.uni-bamberg.de/handle/uniba/47860