Title :
Using multithreshold quadratic sigmoidal neurons to improve classification capability of multilayer perceptrons
Author :
Chiang, Cheng-Chin ; Fu, Hsin-Chia
Author_Institution :
Comput. & Commun. Lab., ITRI, Hsinchu, Taiwan
fDate :
5/1/1994 12:00:00 AM
Abstract :
This letter proposes a new type of neurons called multithreshold quadratic sigmoidal neurons to improve the classification capability of multilayer neural networks. In cooperation with single-threshold quadratic sigmoidal neurons, the multithreshold quadratic sigmoidal neurons can be used to improve the classification capability of multilayer neural networks by a factor of four compared to committee machines and by a factor of two compared to the conventional sigmoidal multilayer perceptrons
Keywords :
feedforward neural nets; pattern recognition; classification capability; multilayer perceptrons; multithreshold quadratic sigmoidal neurons; Computer science; Councils; Feedforward neural networks; Laboratories; Multi-layer neural network; Multilayer perceptrons; Neural networks; Neurons; Nonhomogeneous media; Upper bound;
Journal_Title :
Neural Networks, IEEE Transactions on
