A real-time stepwise supervised learning algorithm for time-series prediction and system identification

Author

Chen, C. L Philip ; Le Clair, S.R.

Author_Institution

Dept. of Comput. Sci. & Eng., Wright State Univ., Dayton, OH, USA

Volume

4

fYear

1996

fDate

3-6 Jun 1996

Firstpage

2009

Abstract

This paper presents a new neural network architecture and a real-time stepwise supervised learning algorithm that rapidly updates the weights of the network while importing new observations. The most significant advantage of the stepwise approach is that the weights of the network can be easily updated so that re-training is not necessary when new data or observations are made available later after the neural network is trained. This feature makes the stepwise updating algorithm perfect for time-series prediction and system identification. The network has also been tested on several data sets and the experimental results are compared with some conventional networks in which more complex architectures and more costly training are needed

Keywords

feedforward neural nets; identification; learning (artificial intelligence); matrix algebra; neural net architecture; prediction theory; real-time systems; time series; autoregression model; linear matrix; multilayer neural networks; neural network architecture; real-time learning; stepwise supervised learning; stepwise updating algorithm; system identification; time-series prediction; Computer architecture; Computer science; MIMO; Neural networks; Nonlinear equations; Prediction algorithms; Real time systems; Supervised learning; System identification; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1996., IEEE International Conference on

Conference_Location

Washington, DC

Print_ISBN

0-7803-3210-5

Type

conf

DOI

10.1109/ICNN.1996.549210

Filename

549210