University of Surrey

Test tubes in the lab Research in the ATI Dance Research

Direct Approaches to Improving the Robustness of Multilayer Neural Networks

Bugmann, G, Sojka, P, Reiss, M, Plumbley, M and Taylor, JG (1992) Direct Approaches to Improving the Robustness of Multilayer Neural Networks In: Artificial Neural Networks, 2: Proceedings of the 1992 International Conference on Artificial Neural Networks (ICANN–92), Brighton, United Kingdom, 4–7 September, 1992, ? - ?.

Full text not available from this repository.


Abstract Multilayer neural networks trained with backpropagation are in general not robust against the loss of a hidden neuron. In this paper we define a form of robustness called 1-node robustness and propose methods to improve it. One approach is based on a modification of the error function by the addition of a "robustness error". It leads to more robust networks but at the cost of a reduced accuracy. A second approach, "pruning-and-duplication", consists of duplicating the neurons whose loss is the most damaging for the network. Pruned neurons are used for the duplication. This procedure leads to robust and accurate networks at low computational cost. It may also prove benefical for generalisation. Both methods are evaluated on the XOR function.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Divisions : Surrey research (other units)
Authors :
Bugmann, G
Sojka, P
Reiss, M
Taylor, JG
Date : 1992
DOI : 10.1016/B978-0-444-89488-5.50049-X
Contributors :
Aleksander, I
Taylor, J
Depositing User : Symplectic Elements
Date Deposited : 17 May 2017 13:36
Last Modified : 23 Jan 2020 18:42

Actions (login required)

View Item View Item


Downloads per month over past year

Information about this web site

© The University of Surrey, Guildford, Surrey, GU2 7XH, United Kingdom.
+44 (0)1483 300800