Pulina, Luca and Tacchella, Armando (2010) An Abstraction-refinement approach to verification of artificial neural networks. In: Computer Aided Verification: 22nd International Conference, CAV 2010, 15-19 July 2010, Edinburgh, United Kingdom. Berlin - Heidelberg, Springer. p. 243-257. (Lecture Notes in Computer Science, 6174/2010). ISBN 978-3-642-14294-9. Conference or Workshop Item.
Full text not available from this repository.
A key problem in the adoption of artificial neural networks in safety-related applications is that misbehaviors can be hardly ruled out with traditional analytical or probabilistic techniques. In this paper we focus on specific networks known as Multi-Layer Perceptrons (MLPs), and we propose a solution to verify their safety using abstractions to Boolean combinations of linear arithmetic constraints. We show that our abstractions are consistent, i.e., whenever the abstract MLP is declared to be safe, the same holds for the concrete one. Spurious counterexamples, on the other hand, trigger refinements and can be leveraged to automate the correction of misbehaviors. We describe an implementation of our approach based on the HySAT solver, detailing the abstraction-refinement process and the automated correction strategy. Finally, we present experimental results confirming the feasibility of our approach on a realistic case study.
I documenti depositati in UnissResearch sono protetti dalle leggi che regolano il diritto d'autore
Repository Staff Only: item control page