Combining time-scale feature extractions with SVMs for stock index forecasting

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

3 Citations (Scopus)

Abstract

Support vector machine (SVM) has appeared as a powerful tool for time series forecasting and demonstrated better performance over other methods. This paper proposes a novel hybrid model which combines time-scale feature extractions with SVM models for stock index forecasting. The time series of explanatory variables are decomposed by the wavelet basis, and the extracted time-scale features then serve as inputs of a SVM model which performs the nonparametric forecasting. Compared with pure SVM models or traditional GARCH models, the performance of the new method is the best. The root-mean-squared forecasting errors are significantly reduced. The results of this study can help investors for controlling and reducing their risks in international investments.

Original languageEnglish
Title of host publicationNeural Information Processing - 13th International Conference, ICONIP 2006, Proceedings
PublisherSpringer Verlag
Pages390-399
Number of pages10
Volume4234 LNCS - III
ISBN (Print)3540464840, 9783540464846
Publication statusPublished - 2006 Jan 1
Event13th International Conference on Neural Information Processing, ICONIP 2006 - Hong Kong, China
Duration: 2006 Oct 32006 Oct 6

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4234 LNCS - III
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other13th International Conference on Neural Information Processing, ICONIP 2006
CountryChina
CityHong Kong
Period06-10-0306-10-06

Fingerprint

Stock Index
Feature Extraction
Forecasting
Feature extraction
Support Vector Machine
Time Scales
Support vector machines
Time series
Time Series Forecasting
GARCH Model
Wavelet Bases
Hybrid Model
Roots
Model

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Huang, S-C., & Wang, H-W. (2006). Combining time-scale feature extractions with SVMs for stock index forecasting. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings (Vol. 4234 LNCS - III, pp. 390-399). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4234 LNCS - III). Springer Verlag.
Huang, Shian-Chang ; Wang, Hsing-Wen. / Combining time-scale feature extractions with SVMs for stock index forecasting. Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Vol. 4234 LNCS - III Springer Verlag, 2006. pp. 390-399 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{c4411bcaa97e4b4b821a54dbeb6a724e,
title = "Combining time-scale feature extractions with SVMs for stock index forecasting",
abstract = "Support vector machine (SVM) has appeared as a powerful tool for time series forecasting and demonstrated better performance over other methods. This paper proposes a novel hybrid model which combines time-scale feature extractions with SVM models for stock index forecasting. The time series of explanatory variables are decomposed by the wavelet basis, and the extracted time-scale features then serve as inputs of a SVM model which performs the nonparametric forecasting. Compared with pure SVM models or traditional GARCH models, the performance of the new method is the best. The root-mean-squared forecasting errors are significantly reduced. The results of this study can help investors for controlling and reducing their risks in international investments.",
author = "Shian-Chang Huang and Hsing-Wen Wang",
year = "2006",
month = "1",
day = "1",
language = "English",
isbn = "3540464840",
volume = "4234 LNCS - III",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "390--399",
booktitle = "Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings",
address = "Germany",

}

Huang, S-C & Wang, H-W 2006, Combining time-scale feature extractions with SVMs for stock index forecasting. in Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. vol. 4234 LNCS - III, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4234 LNCS - III, Springer Verlag, pp. 390-399, 13th International Conference on Neural Information Processing, ICONIP 2006, Hong Kong, China, 06-10-03.

Combining time-scale feature extractions with SVMs for stock index forecasting. / Huang, Shian-Chang; Wang, Hsing-Wen.

Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Vol. 4234 LNCS - III Springer Verlag, 2006. p. 390-399 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4234 LNCS - III).

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

TY - GEN

T1 - Combining time-scale feature extractions with SVMs for stock index forecasting

AU - Huang, Shian-Chang

AU - Wang, Hsing-Wen

PY - 2006/1/1

Y1 - 2006/1/1

N2 - Support vector machine (SVM) has appeared as a powerful tool for time series forecasting and demonstrated better performance over other methods. This paper proposes a novel hybrid model which combines time-scale feature extractions with SVM models for stock index forecasting. The time series of explanatory variables are decomposed by the wavelet basis, and the extracted time-scale features then serve as inputs of a SVM model which performs the nonparametric forecasting. Compared with pure SVM models or traditional GARCH models, the performance of the new method is the best. The root-mean-squared forecasting errors are significantly reduced. The results of this study can help investors for controlling and reducing their risks in international investments.

AB - Support vector machine (SVM) has appeared as a powerful tool for time series forecasting and demonstrated better performance over other methods. This paper proposes a novel hybrid model which combines time-scale feature extractions with SVM models for stock index forecasting. The time series of explanatory variables are decomposed by the wavelet basis, and the extracted time-scale features then serve as inputs of a SVM model which performs the nonparametric forecasting. Compared with pure SVM models or traditional GARCH models, the performance of the new method is the best. The root-mean-squared forecasting errors are significantly reduced. The results of this study can help investors for controlling and reducing their risks in international investments.

UR - http://www.scopus.com/inward/record.url?scp=33750707233&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750707233&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33750707233

SN - 3540464840

SN - 9783540464846

VL - 4234 LNCS - III

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 390

EP - 399

BT - Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings

PB - Springer Verlag

ER -

Huang S-C, Wang H-W. Combining time-scale feature extractions with SVMs for stock index forecasting. In Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Vol. 4234 LNCS - III. Springer Verlag. 2006. p. 390-399. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).