Botoeva, Elena, Kouvaros, Panagiotis, Kronqvist, Jan, Lomuscio, Alessio, Misener, Ruth (2020) Efficient Verification of ReLU-Based Neural Networks via Dependency Analysis. In: Proceedings of the AAAI Conference on Artificial Intelligence. AAAI-20 Technical Tracks , 34 (04). pp. 3291-3299. Association for the Advancement of Artificial Intelligence ISBN 978-1-57735-835-0. (doi:10.1609/aaai.v34i04.5729) (The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided) (KAR id:90815)
The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided. (Contact us about this Publication) | |
Official URL: https://aaai.org/ojs/index.php/AAAI/article/view/5... |
Abstract
We introduce an efficient method for the verification of ReLU-based feed-forward neural networks. We derive an automated procedure that exploits dependency relations between the ReLU nodes, thereby pruning the search tree that needs to be considered by MILP-based formulations of the verification problem. We augment the resulting algorithm with methods for input domain splitting and symbolic interval propagation. We present Venus, the resulting verification toolkit, and evaluate it on the ACAS collision avoidance networks and models trained on the MNIST and CIFAR-10 datasets. The experimental results obtained indicate considerable gains over the present state-of-the-art tools.
Item Type: | Conference or workshop item (Paper) |
---|---|
DOI/Identification number: | 10.1609/aaai.v34i04.5729 |
Subjects: | Q Science > QA Mathematics (inc Computing science) > QA 75 Electronic computers. Computer science |
Divisions: | Divisions > Division of Computing, Engineering and Mathematical Sciences > School of Computing |
Depositing User: | Amy Boaler |
Date Deposited: | 12 Oct 2021 14:08 UTC |
Last Modified: | 18 May 2024 03:16 UTC |
Resource URI: | https://kar.kent.ac.uk/id/eprint/90815 (The current URI for this page, for reference purposes) |
- Export to:
- RefWorks
- EPrints3 XML
- BibTeX
- CSV
- Depositors only (login required):