RoMA: A Method for Neural Network Robustness Measurement and Assessment

Natan Levy*, Guy Katz

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Neural network models have become the leading solution for a large variety of tasks, such as classification, natural language processing, and others. However, their reliability is heavily plagued by adversarial inputs: inputs generated by adding tiny perturbations to correctly-classified inputs, and for which the neural network produces erroneous results. In this paper, we present a new method called Robustness Measurement and Assessment (RoMA), which measures the robustness of a neural network model against such adversarial inputs. Specifically, RoMA determines the probability that a random input perturbation might cause misclassification. The method allows us to provide formal guarantees regarding the expected frequency of errors that a trained model will encounter after deployment. The type of robustness assessment afforded by RoMA is inspired by state-of-the-art certification practices, and could constitute an important step toward integrating neural networks in safety-critical systems.

Original languageEnglish
Title of host publicationNeural Information Processing - 29th International Conference, ICONIP 2022, Proceedings
EditorsMohammad Tanveer, Sonali Agarwal, Seiichi Ozawa, Asif Ekbal, Adam Jatowt
PublisherSpringer Science and Business Media Deutschland GmbH
Pages92-105
Number of pages14
ISBN (Print)9789819916382
DOIs
StatePublished - 2023
Event29th International Conference on Neural Information Processing, ICONIP 2022 - Virtual, Online
Duration: 22 Nov 202226 Nov 2022

Publication series

NameCommunications in Computer and Information Science
Volume1791 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference29th International Conference on Neural Information Processing, ICONIP 2022
CityVirtual, Online
Period22/11/2226/11/22

Bibliographical note

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Keywords

  • Adversarial examples
  • Certification
  • Neural networks
  • Robustness

Fingerprint

Dive into the research topics of 'RoMA: A Method for Neural Network Robustness Measurement and Assessment'. Together they form a unique fingerprint.

Cite this