logo Idiap Research Institute        
 [BibTeX] [Marc21]
The Effects of Optical Thresholding in Backpropagation Neural Networks
Type of publication: Conference paper
Citation: Moerland-95.1
Booktitle: Proceedings of the International Conference on Artificial Neural Networks (ICANN'95 and NeuroNimes'95)
Volume: 2
Year: 1995
Publisher: EC2 & Cie
Location: Paris, France
Organization: ENNS
Address: Paris La Défense, France
ISBN: 2-910085-19-8
Abstract: Sigmoid-like activation functions implemented in analog hardware differ in various ways from the standard sigmoidal function as they are asymmetric, truncated, and have a non-standard gain. It is demonstrated how one can adapt the backpropagation learning rule to compensate for these non-standard sigmoids as available in hardware. This method is applied to multilayer neural networks with all-optical forward propagation and liquid crystal light valves (LCLV) as optical thresholding devices. In this paper the results of software simulations of a backpropagation neural network with five different LCLV activation functions are presented and it is shown that the adapted learning rule performs well with these LCLV curves
Userfields: dates={October 9--13, 1995}, language={English}, ipdmembership={neuron learning},
Projects Idiap
Authors Moerland, Perry
Fiesler, Emile
Saxena, Indu
Editors Fogelman-Soulié, F.
Gallinari, P.
Added by: [UNK]
Total mark: 0