Multi-model integration for long-term time series prediction

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

1 Citation (Scopus)

Abstract

Long-term (multi-step-ahead) time series prediction is a much more challenging task comparing to the short-term (one-step-ahead) time series prediction. This is due to the increasing uncertainty and the lack of knowledge about the future trend. In this paper, we propose a multi-model integration strategy to 1) generate predicted values using multiple predictive models; and then 2) integrate the predicted values to generate a final predicted value as the output. In the first step, a k-nearest-neighbor (k-NN) based least squares support vector machine (LS-SVM) approach is used for long-term time series prediction. An autoregressive model is then employed in the second step to combine the predicted values from the multiple k-NN based LS-SVM models. The proposed multi-model integration strategy is evaluated using six datasets, and the experimental results demonstrate that the proposed strategy consistently outperforms some existing predictors.

Original languageEnglish
Title of host publicationProceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012
Pages116-123
Number of pages8
DOIs
StatePublished - Nov 8 2012
Event2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012 - Las Vegas, NV, United States
Duration: Aug 8 2012Aug 10 2012

Other

Other2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012
CountryUnited States
CityLas Vegas, NV
Period8/8/128/10/12

Fingerprint

Time series
Support vector machines

Keywords

  • autoregressive model
  • k-nearest-neighbor
  • least squares support vector machine (LS-SVM)
  • long-term time series prediction
  • multi-model integration

ASJC Scopus subject areas

  • Information Systems

Cite this

Huang, Z., Shyu, M-L., & Tien, J. M. (2012). Multi-model integration for long-term time series prediction. In Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012 (pp. 116-123). [6302999] https://doi.org/10.1109/IRI.2012.6302999

Multi-model integration for long-term time series prediction. / Huang, Zifang; Shyu, Mei-Ling; Tien, James M.

Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012. 2012. p. 116-123 6302999.

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

Huang, Z, Shyu, M-L & Tien, JM 2012, Multi-model integration for long-term time series prediction. in Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012., 6302999, pp. 116-123, 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012, Las Vegas, NV, United States, 8/8/12. https://doi.org/10.1109/IRI.2012.6302999
Huang Z, Shyu M-L, Tien JM. Multi-model integration for long-term time series prediction. In Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012. 2012. p. 116-123. 6302999 https://doi.org/10.1109/IRI.2012.6302999
Huang, Zifang ; Shyu, Mei-Ling ; Tien, James M. / Multi-model integration for long-term time series prediction. Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012. 2012. pp. 116-123
@inproceedings{2ae9ba4050b2492facc74b92a7ef54fd,
title = "Multi-model integration for long-term time series prediction",
abstract = "Long-term (multi-step-ahead) time series prediction is a much more challenging task comparing to the short-term (one-step-ahead) time series prediction. This is due to the increasing uncertainty and the lack of knowledge about the future trend. In this paper, we propose a multi-model integration strategy to 1) generate predicted values using multiple predictive models; and then 2) integrate the predicted values to generate a final predicted value as the output. In the first step, a k-nearest-neighbor (k-NN) based least squares support vector machine (LS-SVM) approach is used for long-term time series prediction. An autoregressive model is then employed in the second step to combine the predicted values from the multiple k-NN based LS-SVM models. The proposed multi-model integration strategy is evaluated using six datasets, and the experimental results demonstrate that the proposed strategy consistently outperforms some existing predictors.",
keywords = "autoregressive model, k-nearest-neighbor, least squares support vector machine (LS-SVM), long-term time series prediction, multi-model integration",
author = "Zifang Huang and Mei-Ling Shyu and Tien, {James M.}",
year = "2012",
month = "11",
day = "8",
doi = "10.1109/IRI.2012.6302999",
language = "English",
isbn = "9781467322843",
pages = "116--123",
booktitle = "Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012",

}

TY - GEN

T1 - Multi-model integration for long-term time series prediction

AU - Huang, Zifang

AU - Shyu, Mei-Ling

AU - Tien, James M.

PY - 2012/11/8

Y1 - 2012/11/8

N2 - Long-term (multi-step-ahead) time series prediction is a much more challenging task comparing to the short-term (one-step-ahead) time series prediction. This is due to the increasing uncertainty and the lack of knowledge about the future trend. In this paper, we propose a multi-model integration strategy to 1) generate predicted values using multiple predictive models; and then 2) integrate the predicted values to generate a final predicted value as the output. In the first step, a k-nearest-neighbor (k-NN) based least squares support vector machine (LS-SVM) approach is used for long-term time series prediction. An autoregressive model is then employed in the second step to combine the predicted values from the multiple k-NN based LS-SVM models. The proposed multi-model integration strategy is evaluated using six datasets, and the experimental results demonstrate that the proposed strategy consistently outperforms some existing predictors.

AB - Long-term (multi-step-ahead) time series prediction is a much more challenging task comparing to the short-term (one-step-ahead) time series prediction. This is due to the increasing uncertainty and the lack of knowledge about the future trend. In this paper, we propose a multi-model integration strategy to 1) generate predicted values using multiple predictive models; and then 2) integrate the predicted values to generate a final predicted value as the output. In the first step, a k-nearest-neighbor (k-NN) based least squares support vector machine (LS-SVM) approach is used for long-term time series prediction. An autoregressive model is then employed in the second step to combine the predicted values from the multiple k-NN based LS-SVM models. The proposed multi-model integration strategy is evaluated using six datasets, and the experimental results demonstrate that the proposed strategy consistently outperforms some existing predictors.

KW - autoregressive model

KW - k-nearest-neighbor

KW - least squares support vector machine (LS-SVM)

KW - long-term time series prediction

KW - multi-model integration

UR - http://www.scopus.com/inward/record.url?scp=84868310661&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868310661&partnerID=8YFLogxK

U2 - 10.1109/IRI.2012.6302999

DO - 10.1109/IRI.2012.6302999

M3 - Conference contribution

SN - 9781467322843

SP - 116

EP - 123

BT - Proceedings of the 2012 IEEE 13th International Conference on Information Reuse and Integration, IRI 2012

ER -