License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.CONCUR.2019.1
URN: urn:nbn:de:0030-drops-109036
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2019/10903/
Go to the corresponding LIPIcs Volume Portal

Kwiatkowska, Marta Z.

Safety Verification for Deep Neural Networks with Provable Guarantees (Invited Paper)

pdf-format:
LIPIcs-CONCUR-2019-1.pdf (1 MB)

Abstract

Computing systems are becoming ever more complex, increasingly often incorporating deep learning components. Since deep learning is unstable with respect to adversarial perturbations, there is a need for rigorous software development methodologies that encompass machine learning. This paper describes progress with developing automated verification techniques for deep neural networks to ensure safety and robustness of their decisions with respect to input perturbations. This includes novel algorithms based on feature-guided search, games, global optimisation and Bayesian methods.

BibTeX - Entry

@InProceedings{kwiatkowska:LIPIcs:2019:10903,
  author =	{Marta Z. Kwiatkowska},
  title =	{{Safety Verification for Deep Neural Networks with Provable Guarantees (Invited Paper)}},
  booktitle =	{30th International Conference on Concurrency Theory (CONCUR 2019)},
  pages =	{1:1--1:5},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-121-4},
  ISSN =	{1868-8969},
  year =	{2019},
  volume =	{140},
  editor =	{Wan Fokkink and Rob van Glabbeek},
  publisher =	{Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2019/10903},
  URN =		{urn:nbn:de:0030-drops-109036},
  doi =		{10.4230/LIPIcs.CONCUR.2019.1},
  annote =	{Keywords: Neural networks, robustness, formal verification, Bayesian neural networks}
}

Keywords: Neural networks, robustness, formal verification, Bayesian neural networks
Collection: 30th International Conference on Concurrency Theory (CONCUR 2019)
Issue Date: 2019
Date of publication: 20.08.2019

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI