Multi-SVR based fuzzy modeling method for non-stationary time series

Shu Kuan Lin; Li Jia Zhi; Shao Min Zhang; Jian Zhong Qiao; Guo Ren Wang; Ge Yu

Multi-SVR based fuzzy modeling method for non-stationary time series

Shu Kuan Lin^*, Li Jia Zhi, Shao Min Zhang, Jian Zhong Qiao, Guo Ren Wang, Ge Yu

^*Corresponding author for this work

Northeastern University China

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

Abstract

A new approach for modeling non-stationary time series was introduced in this paper. Combine the fuzzy segmentation which was proposed by Janos Abonyi with Support Vector Machines (SVMs). Firstly, a modified Support Vector Regression (SVR) was proposed; Secondly, fuzzy segment information was combined with SVR by heuristic weighting method; Thirdly, we discussed a model based on multi-SVR. Experimental results show that the method proposed in this paper has great practical values for non-stationary time series modeling.

Original language	English
Pages (from-to)	1929-1932
Number of pages	4
Journal	Tien Tzu Hsueh Pao/Acta Electronica Sinica
Volume	34
Issue number	10
Publication status	Published - Oct 2006
Externally published	Yes

Keywords

Fuzzy segmentation
Heuristic weighting method
Multi support vector regression
Non-stationary time series

Cite this

@article{27cff99ecb534efd963f4a3fdcb57f11,

title = "Multi-SVR based fuzzy modeling method for non-stationary time series",

abstract = "A new approach for modeling non-stationary time series was introduced in this paper. Combine the fuzzy segmentation which was proposed by Janos Abonyi with Support Vector Machines (SVMs). Firstly, a modified Support Vector Regression (SVR) was proposed; Secondly, fuzzy segment information was combined with SVR by heuristic weighting method; Thirdly, we discussed a model based on multi-SVR. Experimental results show that the method proposed in this paper has great practical values for non-stationary time series modeling.",

keywords = "Fuzzy segmentation, Heuristic weighting method, Multi support vector regression, Non-stationary time series",

author = "Lin, {Shu Kuan} and Zhi, {Li Jia} and Zhang, {Shao Min} and Qiao, {Jian Zhong} and Wang, {Guo Ren} and Ge Yu",

year = "2006",

month = oct,

language = "English",

volume = "34",

pages = "1929--1932",

journal = "Tien Tzu Hsueh Pao/Acta Electronica Sinica",

issn = "0372-2112",

publisher = "Chinese Institute of Electronics",

number = "10",

}

TY - JOUR

T1 - Multi-SVR based fuzzy modeling method for non-stationary time series

AU - Lin, Shu Kuan

AU - Zhi, Li Jia

AU - Zhang, Shao Min

AU - Qiao, Jian Zhong

AU - Wang, Guo Ren

AU - Yu, Ge

PY - 2006/10

Y1 - 2006/10

N2 - A new approach for modeling non-stationary time series was introduced in this paper. Combine the fuzzy segmentation which was proposed by Janos Abonyi with Support Vector Machines (SVMs). Firstly, a modified Support Vector Regression (SVR) was proposed; Secondly, fuzzy segment information was combined with SVR by heuristic weighting method; Thirdly, we discussed a model based on multi-SVR. Experimental results show that the method proposed in this paper has great practical values for non-stationary time series modeling.

AB - A new approach for modeling non-stationary time series was introduced in this paper. Combine the fuzzy segmentation which was proposed by Janos Abonyi with Support Vector Machines (SVMs). Firstly, a modified Support Vector Regression (SVR) was proposed; Secondly, fuzzy segment information was combined with SVR by heuristic weighting method; Thirdly, we discussed a model based on multi-SVR. Experimental results show that the method proposed in this paper has great practical values for non-stationary time series modeling.

KW - Fuzzy segmentation

KW - Heuristic weighting method

KW - Multi support vector regression

KW - Non-stationary time series

UR - http://www.scopus.com/inward/record.url?scp=33846105059&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:33846105059

SN - 0372-2112

VL - 34

SP - 1929

EP - 1932

JO - Tien Tzu Hsueh Pao/Acta Electronica Sinica

JF - Tien Tzu Hsueh Pao/Acta Electronica Sinica

IS - 10

ER -

Multi-SVR based fuzzy modeling method for non-stationary time series

Abstract

Keywords

Other files and links

Fingerprint

Cite this