Selection of the optimum number of hidden layers in neuro-fuzzy GMDH

Author

Ichihashi, H. ; Harada, N. ; Nagasaka, K.

Author_Institution

Coll. of Eng., Osaka Prefectural Univ., Sakai, Japan

Volume

3

fYear

1995

fDate

20-24 Mar 1995

Firstpage

1519

Abstract

An adaptive learning network (ALN) of group method of data handling (GMDH) type with error backpropagation is proposed, in which Gaussian radial basis functions (RBF) networks are applied to the partial descriptions of the GMDH. Optimum number of hidden layers in the ALN is selected applying the differential minimum bias criterion (DMC) and the Akaike´s information criterion. The validity of these two criteria are confirmed with the cross validation technique and the average mean log-likelihood

Keywords

adaptive systems; feedforward neural nets; fuzzy logic; identification; information theory; learning systems; maximum likelihood estimation; Akaike´s information criterion; Gaussian radial basis functions networks; adaptive learning network; average mean log-likelihood; cross validation; differential minimum bias criterion; group method of data handling; neuro-fuzzy GMDH; optimum number selection; Adaptive systems; Backpropagation; Educational institutions; Fuzzy neural networks; Industrial engineering; Input variables; Intelligent networks; Maximum likelihood estimation; Nonlinear systems; Parameter estimation;

fLanguage

English

Publisher

ieee

Conference_Titel

Fuzzy Systems, 1995. International Joint Conference of the Fourth IEEE International Conference on Fuzzy Systems and The Second International Fuzzy Engineering Symposium., Proceedings of 1995 IEEE Int

Conference_Location

Yokohama

Print_ISBN

0-7803-2461-7

Type

conf

DOI

10.1109/FUZZY.1995.409880

Filename

409880