On-the-fly Black-Box Probably Approximately Correct Checking of Recurrent Neural Networks - Machine Learning and Knowledge Extraction Access content directly
Conference Papers Year : 2020

On-the-fly Black-Box Probably Approximately Correct Checking of Recurrent Neural Networks

Franz Mayr
  • Function : Author
  • PersonId : 1043684
Ramiro Visca
  • Function : Author
  • PersonId : 1115833
Sergio Yovine
  • Function : Author
  • PersonId : 1043685

Abstract

We propose a procedure for checking properties of recurrent neural networks used for language modeling and sequence classification. Our approach is a case of black-box checking based on learning a probably approximately correct, regular approximation of the intersection of the language of the black-box (the network) with the complement of the property to be checked, without explicitly building individual representations of them. When the algorithm returns an empty language, there is a proven upper bound on the probability of the network not verifying the requirement. When the returned language is nonempty, it is certain the network does not satisfy the property. In this case, an explicit and interpretable characterization of the error is output together with sequences of the network truly violating the property. Besides, our approach does not require resorting to an external decision procedure for verification nor fixing a specific property specification formalism.
Fichier principal
Vignette du fichier
497121_1_En_19_Chapter.pdf (576.37 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03414742 , version 1 (04-11-2021)

Licence

Attribution

Identifiers

Cite

Franz Mayr, Ramiro Visca, Sergio Yovine. On-the-fly Black-Box Probably Approximately Correct Checking of Recurrent Neural Networks. 4th International Cross-Domain Conference for Machine Learning and Knowledge Extraction (CD-MAKE), Aug 2020, Dublin, Ireland. pp.343-363, ⟨10.1007/978-3-030-57321-8_19⟩. ⟨hal-03414742⟩
17 View
24 Download

Altmetric

Share

Gmail Facebook X LinkedIn More