DocumentCode :
2884831
Title :
A new neural network architecture based on quadratic function neurons
Author :
Leung, Chung S. ; Cheung, K.F. ; Poon, M.C.
Author_Institution :
Dept. of Inf. Eng., Chinese Univ. of Hong Kong, Hong Kong
fYear :
1991
fDate :
16-17 Jun 1991
Firstpage :
264
Abstract :
In this paper, a class of multilayer perceptrons known as rotational quadratic function neural networks (RQFNN) is introduced. The rotational quadratic function neuron (RQFN), at the center of this class of networks, is a particular implementation of the quadratic function neuron (QFN). Compared with the traditional implementation, the RQFN requires much less fan-ins and thus much smaller cross-connection volume. The economy of the fan-ins and the cross connection volumes facilitates the mapping of the model onto silicon
Keywords :
learning systems; neural nets; back propagation model; constrained type learning; cross-connection volume; fan-ins reduction; neural network architecture; quadratic function neurons; rotational quadratic function; training; Ambient intelligence; Backpropagation; Cities and towns; Lungs; Multi-layer neural network; Multilayer perceptrons; Neural networks; Neurons; Prototypes; Silicon;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Circuits and Systems, 1991. Conference Proceedings, China., 1991 International Conference on
Conference_Location :
Shenzhen
Type :
conf
DOI :
10.1109/CICCAS.1991.184335
Filename :
184335
Link To Document :
بازگشت