Title :
Configuring RBF neural networks
Author :
Sohn, Illsoo ; Ansari, Nayeem
Author_Institution :
Dept. of Electr. & Comput. Eng., New Jersey Inst. of Technol., Newark, NJ
fDate :
4/2/1998 12:00:00 AM
Abstract :
A novel method (based on the characteristics of scatter matrices and frequency-sensitive competitive learning) for training the hidden layer of a radial basis function neural network is proposed. The method is demonstrated to be robust and to outperform the state-of-the-art algorithm
Keywords :
S-matrix theory; feedforward neural nets; unsupervised learning; RBF neural network; SBC; frequency-sensitive competitive learning; hidden layer; radial basis function neural network; robust algorithm; scatter matrix; training;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19980469
