Continuous verification of machine learning: a declarative programming approach

Komendantskaya, Ekaterina, Kokke, Wen and Kienitz, Daniel (2020) Continuous verification of machine learning: a declarative programming approach. In Proceedings of the 22nd International Symposium on Principles and Practice of Declarative Programming, PPDP 2020 - Part of BOPL 2020 - Bologna Federated Conference on Programming Languages 2020. Association for Computing Machinery.. (doi:10.1145/3414080.3414081).

Record type: Conference or Workshop Item (Paper)

Abstract

In this invited talk, we discuss state of the art in neural network verification. We propose the term continuous verification to characterise the family of methods that explore continuous nature of machine learning algorithms. We argue that methods of continuous verification must rely on robust programming language infrastructure (refinement types, automated proving, type-driven program synthesis), which provides a major opportunity for the declarative programming language community. Keywords: Neural Networks, Verification, AI.

This record has no associated files available for download.

More information

Published date: 8 September 2020

Additional Information: Funding Information: We acknowledge support of the UK National Cyber Security Center grant SecCon-NN: Neural Networks with Security Contracts - towards lightweight, modular security for neural networks and the UK Research Institute in Verified Trustworthy Software Systems (VETSS)-funded research project CONVENER: Continuous Verification of Neural Networks. Publisher Copyright: © 2020 Owner/Author.

Venue - Dates: 22nd International Symposium on Principles and Practice of Declarative Programming, PPDP 2020 - Part of 2020 Bologna Federated Conference on Programming Languages, BOPL 2020, , Bologna, Online, Italy, 2020-09-08 - 2020-09-10

Learn more about Cyber Physical Systems research