Combining wavelet-based feature extractions with relevance vector machines for stock index forecasting

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

The relevance vector machine (RVM) is a Bayesian version of the support vector machine, which with a sparse model representation has appeared to be a powerful tool for time-series forecasting. The RVM has demonstrated better performance over other methods such as neural networks or autoregressive integrated moving average based models. This study proposes a hybrid model that combines wavelet-based feature extractions with RVM models to forecast stock indices. The time series of explanatory variables are decomposed using some wavelet bases and the extracted time-scale features serve as inputs of an RVM to perform the non-parametric regression and forecasting. Compared with traditional forecasting models, our proposed method performs best. The root-mean-squared forecasting errors are significantly reduced.

Original languageEnglish
Pages (from-to)133-149
Number of pages17
JournalExpert Systems
Volume25
Issue number2
DOIs
Publication statusPublished - 2008 May 1

Fingerprint

Relevance Vector Machine
Stock Index
Feature Extraction
Forecasting
Feature extraction
Wavelets
Time series
Time Series Forecasting
Wavelet Bases
Moving Average
Hybrid Model
Nonparametric Regression
Model
Forecast
Support Vector Machine
Time Scales
Roots
Support vector machines
Neural Networks
Neural networks

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Theoretical Computer Science
  • Computational Theory and Mathematics
  • Artificial Intelligence

Cite this

@article{0db2364f5a24496c85168397ccd623cc,
title = "Combining wavelet-based feature extractions with relevance vector machines for stock index forecasting",
abstract = "The relevance vector machine (RVM) is a Bayesian version of the support vector machine, which with a sparse model representation has appeared to be a powerful tool for time-series forecasting. The RVM has demonstrated better performance over other methods such as neural networks or autoregressive integrated moving average based models. This study proposes a hybrid model that combines wavelet-based feature extractions with RVM models to forecast stock indices. The time series of explanatory variables are decomposed using some wavelet bases and the extracted time-scale features serve as inputs of an RVM to perform the non-parametric regression and forecasting. Compared with traditional forecasting models, our proposed method performs best. The root-mean-squared forecasting errors are significantly reduced.",
author = "Shian-Chang Huang and Tung-Kuang Wu",
year = "2008",
month = "5",
day = "1",
doi = "10.1111/j.1468-0394.2008.00443.x",
language = "English",
volume = "25",
pages = "133--149",
journal = "Expert Systems",
issn = "0266-4720",
publisher = "Wiley-Blackwell",
number = "2",

}

Combining wavelet-based feature extractions with relevance vector machines for stock index forecasting. / Huang, Shian-Chang; Wu, Tung-Kuang.

In: Expert Systems, Vol. 25, No. 2, 01.05.2008, p. 133-149.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Combining wavelet-based feature extractions with relevance vector machines for stock index forecasting

AU - Huang, Shian-Chang

AU - Wu, Tung-Kuang

PY - 2008/5/1

Y1 - 2008/5/1

N2 - The relevance vector machine (RVM) is a Bayesian version of the support vector machine, which with a sparse model representation has appeared to be a powerful tool for time-series forecasting. The RVM has demonstrated better performance over other methods such as neural networks or autoregressive integrated moving average based models. This study proposes a hybrid model that combines wavelet-based feature extractions with RVM models to forecast stock indices. The time series of explanatory variables are decomposed using some wavelet bases and the extracted time-scale features serve as inputs of an RVM to perform the non-parametric regression and forecasting. Compared with traditional forecasting models, our proposed method performs best. The root-mean-squared forecasting errors are significantly reduced.

AB - The relevance vector machine (RVM) is a Bayesian version of the support vector machine, which with a sparse model representation has appeared to be a powerful tool for time-series forecasting. The RVM has demonstrated better performance over other methods such as neural networks or autoregressive integrated moving average based models. This study proposes a hybrid model that combines wavelet-based feature extractions with RVM models to forecast stock indices. The time series of explanatory variables are decomposed using some wavelet bases and the extracted time-scale features serve as inputs of an RVM to perform the non-parametric regression and forecasting. Compared with traditional forecasting models, our proposed method performs best. The root-mean-squared forecasting errors are significantly reduced.

UR - http://www.scopus.com/inward/record.url?scp=42949171426&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=42949171426&partnerID=8YFLogxK

U2 - 10.1111/j.1468-0394.2008.00443.x

DO - 10.1111/j.1468-0394.2008.00443.x

M3 - Article

AN - SCOPUS:42949171426

VL - 25

SP - 133

EP - 149

JO - Expert Systems

JF - Expert Systems

SN - 0266-4720

IS - 2

ER -