The paper addresses the analysis of robustness over training time issue. Robustness is evaluated in the large, without assuming the small perturbation hypothesis, by means of randomised algorithms. We discovered that robustness is a strict property of the model -as it is accuracy- and, hence, it depends on the particular neural network family, application, training algorithm and training starting point. Complex neural networks are hence not necessarily more robust than less complex topologies. An early stopping algorithm is finally suggested which extends the one based on the test set inspection with robustness aspects.
A training-time analysis of robustness in feed-forward neural networks / C. Alippi, D. Sana, F. Scotti - In: Proceedings of 2004 IEEE International joint conference on neural networks : 25-29 july 2004. 4Piscataway : Institute of electrical and electronics engineers, 2004. - ISBN 0780383591. - pp. 2853-2858 (( convegno IEEE International Joint Conference on Neural Networks nel 2004.
A training-time analysis of robustness in feed-forward neural networks
D. SanaSecondo
;F. ScottiUltimo
2004
Abstract
The paper addresses the analysis of robustness over training time issue. Robustness is evaluated in the large, without assuming the small perturbation hypothesis, by means of randomised algorithms. We discovered that robustness is a strict property of the model -as it is accuracy- and, hence, it depends on the particular neural network family, application, training algorithm and training starting point. Complex neural networks are hence not necessarily more robust than less complex topologies. An early stopping algorithm is finally suggested which extends the one based on the test set inspection with robustness aspects.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.