Faster and better training of multi-layer perceptron for forecasting problems

Author

Laddad, R.R. ; Desai, U.B. ; Poonacha, P.G.

Author_Institution

Dept. of Electr. Eng., Indian Inst. of Technol., Bombay, India

fYear

1994

fDate

6-8 Sep 1994

Firstpage

88

Lastpage

97

Abstract

New methods for training multi-layer perceptron network for forecasting problems are presented. The first method exploits spectral characteristics of time series to get faster learning and improved prediction accuracy. A neural network scheme for real time implementation of this method is also presented. The second method suggests the use of two new weight initialization schemes which give very fast convergence besides giving better prediction. The foreign exchange time series is used to illustrate the efficacy of the proposed methods

Keywords

convergence; forecasting theory; foreign exchange trading; learning (artificial intelligence); multilayer perceptrons; spectral analysis; time series; fast convergence; forecasting problems; foreign exchange; learning; multilayer perceptron; neural network scheme; prediction accuracy; spectral characteristics; time series; weight initialization schemes; Accuracy; Backpropagation algorithms; Convergence; Delay effects; Electronic mail; Load forecasting; Multilayer perceptrons; Neural networks; Noise figure; Technology forecasting;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks for Signal Processing [1994] IV. Proceedings of the 1994 IEEE Workshop

Conference_Location

Ermioni

Print_ISBN

0-7803-2026-3

Type

conf

DOI

10.1109/NNSP.1994.366060

Filename

366060