On performance measures of artificial neural networks trained by structural learning algorithms

Author

Kozma, R. ; Kitamura, M. ; Malinowski, A. ; Zurada, J.M.

Author_Institution

Dept. of Nucl. Eng., Tohoku Univ., Sendai, Japan

fYear

1995

Firstpage

22

Lastpage

25

Abstract

Structural learning in multi layer, feedforward neural networks was studied using M. Ishikawa´s (1994) modified backpropagation algorithm with forgetting of the connection weights. The proper choice of forgetting constant was investigated previously but no generally accepted method has been established yet. The generalization rate of the trained network is analyzed as a possible means of selecting optimum model parameters. The results are illustrated using R.A. Fisher´s (1936) IRIS data and anomaly detection in time series

Keywords

backpropagation; feedforward neural nets; generalisation (artificial intelligence); multilayer perceptrons; software performance evaluation; IRIS data; anomaly detection; artificial neural networks; connection weights; forgetting constant; generalization rate; modified backpropagation algorithm; multi layer feedforward neural networks; optimum model parameters; performance measures; structural learning algorithms; time series; Artificial neural networks; Computer networks; Error analysis; Iris; Multi-layer neural network; Neural networks; Signal analysis; Signal processing; Testing; Time series analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Artificial Neural Networks and Expert Systems, 1995. Proceedings., Second New Zealand International Two-Stream Conference on

Conference_Location

Dunedin

Print_ISBN

0-8186-7174-2

Type

conf

DOI

10.1109/ANNES.1995.499430

Filename

499430