Gaussian process for nonstationary time series prediction

Sofiane Brahim-Belhouari, Amine Bermak

Research output: Contribution to journalArticle

142 Citations (Scopus)

Abstract

In this paper, the problem of time series prediction is studied. A Bayesian procedure based on Gaussian process models using a nonstationary covariance function is proposed. Experiments proved the approach effectiveness with an excellent prediction and a good tracking. The conceptual simplicity, and good performance of Gaussian process models should make them very attractive for a wide range of problems.

Original languageEnglish
Pages (from-to)705-712
Number of pages8
JournalComputational Statistics and Data Analysis
Volume47
Issue number4
DOIs
Publication statusPublished - 1 Nov 2004
Externally publishedYes

Fingerprint

Non-stationary Time Series
Time Series Prediction
Gaussian Model
Gaussian Process
Process Model
Time series
Covariance Function
Simplicity
Prediction
Range of data
Experiment
Experiments

Keywords

  • Bayesian learning
  • Gaussian processes
  • Prediction theory
  • Time series

ASJC Scopus subject areas

  • Statistics and Probability
  • Computational Theory and Mathematics
  • Computational Mathematics
  • Applied Mathematics

Cite this

Gaussian process for nonstationary time series prediction. / Brahim-Belhouari, Sofiane; Bermak, Amine.

In: Computational Statistics and Data Analysis, Vol. 47, No. 4, 01.11.2004, p. 705-712.

Research output: Contribution to journalArticle

@article{66ac056d2d6a497782a0624aa65016b1,
title = "Gaussian process for nonstationary time series prediction",
abstract = "In this paper, the problem of time series prediction is studied. A Bayesian procedure based on Gaussian process models using a nonstationary covariance function is proposed. Experiments proved the approach effectiveness with an excellent prediction and a good tracking. The conceptual simplicity, and good performance of Gaussian process models should make them very attractive for a wide range of problems.",
keywords = "Bayesian learning, Gaussian processes, Prediction theory, Time series",
author = "Sofiane Brahim-Belhouari and Amine Bermak",
year = "2004",
month = "11",
day = "1",
doi = "10.1016/j.csda.2004.02.006",
language = "English",
volume = "47",
pages = "705--712",
journal = "Computational Statistics and Data Analysis",
issn = "0167-9473",
publisher = "Elsevier",
number = "4",

}

TY - JOUR

T1 - Gaussian process for nonstationary time series prediction

AU - Brahim-Belhouari, Sofiane

AU - Bermak, Amine

PY - 2004/11/1

Y1 - 2004/11/1

N2 - In this paper, the problem of time series prediction is studied. A Bayesian procedure based on Gaussian process models using a nonstationary covariance function is proposed. Experiments proved the approach effectiveness with an excellent prediction and a good tracking. The conceptual simplicity, and good performance of Gaussian process models should make them very attractive for a wide range of problems.

AB - In this paper, the problem of time series prediction is studied. A Bayesian procedure based on Gaussian process models using a nonstationary covariance function is proposed. Experiments proved the approach effectiveness with an excellent prediction and a good tracking. The conceptual simplicity, and good performance of Gaussian process models should make them very attractive for a wide range of problems.

KW - Bayesian learning

KW - Gaussian processes

KW - Prediction theory

KW - Time series

UR - http://www.scopus.com/inward/record.url?scp=8344270238&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=8344270238&partnerID=8YFLogxK

U2 - 10.1016/j.csda.2004.02.006

DO - 10.1016/j.csda.2004.02.006

M3 - Article

AN - SCOPUS:8344270238

VL - 47

SP - 705

EP - 712

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

SN - 0167-9473

IS - 4

ER -