Please use this identifier to cite or link to this item:
http://hdl.handle.net/10397/711
Title: | Minimum description length neural networks for time series prediction | Authors: | Small, M Tse, CKM |
Issue Date: | Dec-2002 | Source: | Physical review. E, Statistical, nonlinear, and soft matter physics, Dec. 2002, v. 66, no. 6, 066701, p. 1-12 | Abstract: | Artificial neural networks (ANN) are typically composed of a large number of nonlinear functions (neurons) each with several linear and nonlinear parameters that are fitted to data through a computationally intensive training process. Longer training results in a closer fit to the data, but excessive training will lead to overfitting. We propose an alternative scheme that has previously been described for radial basis functions (RBF). We show that fundamental differences between ANN and RBF make application of this scheme to ANN nontrivial. Under this scheme, the training process is replaced by an optimal fitting routine, and overfitting is avoided by controlling the number of neurons in the network. We show that for time series modeling and prediction, this procedure leads to small models (few neurons) that mimic the underlying dynamics of the system well and do not overfit the data. We apply this algorithm to several computational and real systems including chaotic differential equations, the annual sunspot count, and experimental data obtained from a chaotic laser. Our experiments indicate that the structural differences between ANN and RBF make ANN particularly well suited to modeling chaotic time series data. | Keywords: | Algorithms Computer simulation Data reduction Extrapolation Information analysis Iterative methods Polynomials Radial basis function networks Real time systems Time series analysis |
Publisher: | American Physical Society | Journal: | Physical review. E, Statistical, nonlinear, and soft matter physics | ISSN: | 1539-3755 | EISSN: | 1550-2376 | DOI: | 10.1103/PhysRevE.66.066701 | Rights: | Copyright 2002 by the American Physical Society |
Appears in Collections: | Journal/Magazine Article |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
series-prediction_02.pdf | 150.57 kB | Adobe PDF | View/Open |
Page views
125
Last Week
1
1
Last month
Citations as of Apr 14, 2024
Downloads
300
Citations as of Apr 14, 2024
SCOPUSTM
Citations
74
Last Week
0
0
Last month
0
0
Citations as of Apr 5, 2024
WEB OF SCIENCETM
Citations
66
Last Week
0
0
Last month
0
0
Citations as of Apr 18, 2024
Google ScholarTM
Check
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.