On neuro-wavelet modeling

Fionn Murtagh, Jean Luc Starck, Olivier Renuad

Research output: Contribution to journalReview article

92 Citations (Scopus)

Abstract

We survey a number of applications of the wavelet transform in time series prediction. The Haar à trous wavelet transform is proposed as a means of handling time series data when future data is unknown. Results are exemplified on financial futures and S&P500 data. Nonlinear and linear multiresolution autoregressionmodels are studied. Experimentally, we show that multiresolution approaches can outperform the traditional single resolution approach to modeling and prediction.

Original languageEnglish
Pages (from-to)475-484
Number of pages10
JournalDecision Support Systems
Volume37
Issue number4
Early online date3 Jul 2003
DOIs
Publication statusPublished - 1 Sep 2004
Externally publishedYes

Fingerprint

Wavelet Analysis
Wavelet transforms
Time series
Prediction
Wavelets
Wavelet transform
Modeling
Surveys and Questionnaires
Time series data

Cite this

Murtagh, Fionn ; Starck, Jean Luc ; Renuad, Olivier. / On neuro-wavelet modeling. In: Decision Support Systems. 2004 ; Vol. 37, No. 4. pp. 475-484.
@article{05dd487d88554a5095a68661951032bd,
title = "On neuro-wavelet modeling",
abstract = "We survey a number of applications of the wavelet transform in time series prediction. The Haar {\`a} trous wavelet transform is proposed as a means of handling time series data when future data is unknown. Results are exemplified on financial futures and S&P500 data. Nonlinear and linear multiresolution autoregressionmodels are studied. Experimentally, we show that multiresolution approaches can outperform the traditional single resolution approach to modeling and prediction.",
keywords = "{\`A} trous wavelet transform, Feature selection, Haar wavelet transform, Time series forecasting",
author = "Fionn Murtagh and Starck, {Jean Luc} and Olivier Renuad",
year = "2004",
month = "9",
day = "1",
doi = "10.1016/S0167-9236(03)00092-7",
language = "English",
volume = "37",
pages = "475--484",
journal = "Decision Support Systems",
issn = "0167-9236",
publisher = "Elsevier",
number = "4",

}

On neuro-wavelet modeling. / Murtagh, Fionn; Starck, Jean Luc; Renuad, Olivier.

In: Decision Support Systems, Vol. 37, No. 4, 01.09.2004, p. 475-484.

Research output: Contribution to journalReview article

TY - JOUR

T1 - On neuro-wavelet modeling

AU - Murtagh, Fionn

AU - Starck, Jean Luc

AU - Renuad, Olivier

PY - 2004/9/1

Y1 - 2004/9/1

N2 - We survey a number of applications of the wavelet transform in time series prediction. The Haar à trous wavelet transform is proposed as a means of handling time series data when future data is unknown. Results are exemplified on financial futures and S&P500 data. Nonlinear and linear multiresolution autoregressionmodels are studied. Experimentally, we show that multiresolution approaches can outperform the traditional single resolution approach to modeling and prediction.

AB - We survey a number of applications of the wavelet transform in time series prediction. The Haar à trous wavelet transform is proposed as a means of handling time series data when future data is unknown. Results are exemplified on financial futures and S&P500 data. Nonlinear and linear multiresolution autoregressionmodels are studied. Experimentally, we show that multiresolution approaches can outperform the traditional single resolution approach to modeling and prediction.

KW - À trous wavelet transform

KW - Feature selection

KW - Haar wavelet transform

KW - Time series forecasting

UR - http://www.scopus.com/inward/record.url?scp=2442717994&partnerID=8YFLogxK

U2 - 10.1016/S0167-9236(03)00092-7

DO - 10.1016/S0167-9236(03)00092-7

M3 - Review article

VL - 37

SP - 475

EP - 484

JO - Decision Support Systems

JF - Decision Support Systems

SN - 0167-9236

IS - 4

ER -