Title :
Optimization of activation functions in multilayer neural network applied to pattern classification
Author :
Nakayama, Kenji ; KIMURA, Yoshinori
Author_Institution :
Dept. of Electr. & Comput. Eng., Kanazawa Univ., Japan
fDate :
27 Jun-2 Jul 1994
Abstract :
An optimization method of activation functions is proposed. Three typical functions are combined in hidden layers. Contribution of the functions is evaluated using three criteria. The useful functions, are selected or multiplied in the learning process. Problems of parity and of counting `1´ in bit-patterns can be solved by the proposed method with the suitable functions and the minimum number of hidden units
Keywords :
backpropagation; multilayer perceptrons; optimisation; pattern classification; transfer functions; activation functions; counting; hidden layers; multilayer neural network; optimization method; parity; pattern classification; Backpropagation algorithms; Computer simulation; Design optimization; Electronic mail; Equations; Intelligent networks; Multi-layer neural network; Neural networks; Optimization methods; Pattern classification;
Conference_Titel :
Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-1901-X
DOI :
10.1109/ICNN.1994.374201
