A hybrid predictor for time series prediction

Yen Ping Chen, Sheng Nan Wu, Jeen-Shing Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

This paper presents a hybrid predictor for the CATS (Competition on Artificial Time Series) benchmark. Considering the time series as a sum of two components: the major trend and a residual series, we tackled the prediction problem by a hybrid predictor consisting of two models - a kernel regression model and a recurrent neuro-fuzzy model. The kernel regression model based on Gaussian function expansions was first applied to predict the major trend of the time series. The time series was sectioned into several data sets to obtain the best-fitting regression model. Subsequently, the recurrent neuro-fuzzy model associated with a learning algorithm was used to predict the dynamics of the residual series. The learning algorithm has been developed to construct a minimum size of the recurrent model in state-space representation. The best prediction results were presented and discussed.

Original languageEnglish
Title of host publication2004 IEEE International Joint Conference on Neural Networks - Proceedings
Pages1597-1602
Number of pages6
Volume2
DOIs
Publication statusPublished - 2004
Event2004 IEEE International Joint Conference on Neural Networks - Proceedings - Budapest, Hungary
Duration: 2004 Jul 252004 Jul 29

Other

Other2004 IEEE International Joint Conference on Neural Networks - Proceedings
CountryHungary
CityBudapest
Period04-07-2504-07-29

Fingerprint

Time series
Learning algorithms

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Chen, Y. P., Wu, S. N., & Wang, J-S. (2004). A hybrid predictor for time series prediction. In 2004 IEEE International Joint Conference on Neural Networks - Proceedings (Vol. 2, pp. 1597-1602) https://doi.org/10.1109/IJCNN.2004.1380196
Chen, Yen Ping ; Wu, Sheng Nan ; Wang, Jeen-Shing. / A hybrid predictor for time series prediction. 2004 IEEE International Joint Conference on Neural Networks - Proceedings. Vol. 2 2004. pp. 1597-1602
@inproceedings{1e9ba5bb630744c182b49b6e29b2bac6,
title = "A hybrid predictor for time series prediction",
abstract = "This paper presents a hybrid predictor for the CATS (Competition on Artificial Time Series) benchmark. Considering the time series as a sum of two components: the major trend and a residual series, we tackled the prediction problem by a hybrid predictor consisting of two models - a kernel regression model and a recurrent neuro-fuzzy model. The kernel regression model based on Gaussian function expansions was first applied to predict the major trend of the time series. The time series was sectioned into several data sets to obtain the best-fitting regression model. Subsequently, the recurrent neuro-fuzzy model associated with a learning algorithm was used to predict the dynamics of the residual series. The learning algorithm has been developed to construct a minimum size of the recurrent model in state-space representation. The best prediction results were presented and discussed.",
author = "Chen, {Yen Ping} and Wu, {Sheng Nan} and Jeen-Shing Wang",
year = "2004",
doi = "10.1109/IJCNN.2004.1380196",
language = "English",
isbn = "0780383591",
volume = "2",
pages = "1597--1602",
booktitle = "2004 IEEE International Joint Conference on Neural Networks - Proceedings",

}

Chen, YP, Wu, SN & Wang, J-S 2004, A hybrid predictor for time series prediction. in 2004 IEEE International Joint Conference on Neural Networks - Proceedings. vol. 2, pp. 1597-1602, 2004 IEEE International Joint Conference on Neural Networks - Proceedings, Budapest, Hungary, 04-07-25. https://doi.org/10.1109/IJCNN.2004.1380196

A hybrid predictor for time series prediction. / Chen, Yen Ping; Wu, Sheng Nan; Wang, Jeen-Shing.

2004 IEEE International Joint Conference on Neural Networks - Proceedings. Vol. 2 2004. p. 1597-1602.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A hybrid predictor for time series prediction

AU - Chen, Yen Ping

AU - Wu, Sheng Nan

AU - Wang, Jeen-Shing

PY - 2004

Y1 - 2004

N2 - This paper presents a hybrid predictor for the CATS (Competition on Artificial Time Series) benchmark. Considering the time series as a sum of two components: the major trend and a residual series, we tackled the prediction problem by a hybrid predictor consisting of two models - a kernel regression model and a recurrent neuro-fuzzy model. The kernel regression model based on Gaussian function expansions was first applied to predict the major trend of the time series. The time series was sectioned into several data sets to obtain the best-fitting regression model. Subsequently, the recurrent neuro-fuzzy model associated with a learning algorithm was used to predict the dynamics of the residual series. The learning algorithm has been developed to construct a minimum size of the recurrent model in state-space representation. The best prediction results were presented and discussed.

AB - This paper presents a hybrid predictor for the CATS (Competition on Artificial Time Series) benchmark. Considering the time series as a sum of two components: the major trend and a residual series, we tackled the prediction problem by a hybrid predictor consisting of two models - a kernel regression model and a recurrent neuro-fuzzy model. The kernel regression model based on Gaussian function expansions was first applied to predict the major trend of the time series. The time series was sectioned into several data sets to obtain the best-fitting regression model. Subsequently, the recurrent neuro-fuzzy model associated with a learning algorithm was used to predict the dynamics of the residual series. The learning algorithm has been developed to construct a minimum size of the recurrent model in state-space representation. The best prediction results were presented and discussed.

UR - http://www.scopus.com/inward/record.url?scp=10944228116&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10944228116&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2004.1380196

DO - 10.1109/IJCNN.2004.1380196

M3 - Conference contribution

SN - 0780383591

VL - 2

SP - 1597

EP - 1602

BT - 2004 IEEE International Joint Conference on Neural Networks - Proceedings

ER -

Chen YP, Wu SN, Wang J-S. A hybrid predictor for time series prediction. In 2004 IEEE International Joint Conference on Neural Networks - Proceedings. Vol. 2. 2004. p. 1597-1602 https://doi.org/10.1109/IJCNN.2004.1380196