Active Monitoring of Neural Networks

Active Monitoring of Neural Networks

Lukina, A. (TU Delft Algorithmics)
Schilling, Christian (Universität Konstanz)
Henzinger, Thomas A. (Institute of Science and Technology (IST Austria))

Leiva, Edit Luis A. (editor)
Pruski, Cédric (editor)
Markovich, Réka (editor)
Najjar, Amro (editor)
Schommer, Christoph (editor)

2021

Neural-network classifiers are trained to achieve high prediction accuracy. However, their performance still suffers from frequently appearing inputs of unknown classes. As a component of a cyber-physical system, the classifier in this case can no longer be reliable and is typically retrained. We propose an algorithmic framework for monitoring reliability of a neural network. In contrast to static detection, a monitor wrapped in our framework operates in parallel with the classifier, communicates interpretable labeling queries to the human user, and incrementally adapts to their feedback.

monitoring
neural networks
novelty detection

http://resolver.tudelft.nl/uuid:f468eb51-88f2-4f21-b928-b9516e93e946

BNAIC/BeneLearn 2021: 33rd Benelux Conference on Artificial Intelligence and 30th Belgian-Dutch Conference on Machine Learning

33rd Benelux Conference on Artificial Intelligence and
30th Belgian-Dutch Conference on Machine Learning, 2021-11-10 → 2021-11-12, Esch-sur-Alzette, Luxembourg

Institutional Repository

conference paper

© 2021 A. Lukina, Christian Schilling, Thomas A. Henzinger