Automatic time series analysis for electric load forecasting via support vector regression

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Automatic time series analysis for electric load forecasting via support vector regression. / Maldonado, S.; González, A.; Crone, S.
In: Applied Soft Computing Journal, Vol. 83, 105616, 01.10.2019.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Maldonado, S, González, A & Crone, S 2019, 'Automatic time series analysis for electric load forecasting via support vector regression', Applied Soft Computing Journal, vol. 83, 105616. https://doi.org/10.1016/j.asoc.2019.105616

APA

Maldonado, S., González, A., & Crone, S. (2019). Automatic time series analysis for electric load forecasting via support vector regression. Applied Soft Computing Journal, 83, Article 105616. https://doi.org/10.1016/j.asoc.2019.105616

Vancouver

Maldonado S, González A, Crone S. Automatic time series analysis for electric load forecasting via support vector regression. Applied Soft Computing Journal. 2019 Oct 1;83:105616. Epub 2019 Jul 9. doi: 10.1016/j.asoc.2019.105616

Author

Maldonado, S. ; González, A. ; Crone, S. / Automatic time series analysis for electric load forecasting via support vector regression. In: Applied Soft Computing Journal. 2019 ; Vol. 83.

Bibtex

@article{d6c9800f456148af830b6a369e2ffc32,

title = "Automatic time series analysis for electric load forecasting via support vector regression",

abstract = "In this work, a strategy for automatic lag selection in time series analysis is proposed. The method extends the ideas of feature selection with support vector regression, a powerful machine learning tool that can identify nonlinear patterns effectively thanks to the introduction of a kernel function. The proposed approach follows a backward variable elimination procedure based on gradient descent optimisation, iteratively adjusting the widths of an anisotropic Gaussian kernel. Experiments on four electricity demand forecasting datasets demonstrate the virtues of the proposed approach in terms of predictive performance and correct identification of relevant lags and seasonal patterns, compared to well-known strategies for time series analysis designed for energy load forecasting and state-of-the-art strategies for automatic model selection.",

keywords = "Analytics, Feature selection, Support vector regression, Time series analysis, Electric load forecasting",

author = "S. Maldonado and A. Gonz{\'a}lez and S. Crone",

year = "2019",

month = oct,

day = "1",

doi = "10.1016/j.asoc.2019.105616",

language = "English",

volume = "83",

journal = "Applied Soft Computing Journal",

issn = "1568-4946",

publisher = "Elsevier Science B.V.",

}

RIS

TY - JOUR

T1 - Automatic time series analysis for electric load forecasting via support vector regression

AU - Maldonado, S.

AU - González, A.

AU - Crone, S.

PY - 2019/10/1

Y1 - 2019/10/1

N2 - In this work, a strategy for automatic lag selection in time series analysis is proposed. The method extends the ideas of feature selection with support vector regression, a powerful machine learning tool that can identify nonlinear patterns effectively thanks to the introduction of a kernel function. The proposed approach follows a backward variable elimination procedure based on gradient descent optimisation, iteratively adjusting the widths of an anisotropic Gaussian kernel. Experiments on four electricity demand forecasting datasets demonstrate the virtues of the proposed approach in terms of predictive performance and correct identification of relevant lags and seasonal patterns, compared to well-known strategies for time series analysis designed for energy load forecasting and state-of-the-art strategies for automatic model selection.

AB - In this work, a strategy for automatic lag selection in time series analysis is proposed. The method extends the ideas of feature selection with support vector regression, a powerful machine learning tool that can identify nonlinear patterns effectively thanks to the introduction of a kernel function. The proposed approach follows a backward variable elimination procedure based on gradient descent optimisation, iteratively adjusting the widths of an anisotropic Gaussian kernel. Experiments on four electricity demand forecasting datasets demonstrate the virtues of the proposed approach in terms of predictive performance and correct identification of relevant lags and seasonal patterns, compared to well-known strategies for time series analysis designed for energy load forecasting and state-of-the-art strategies for automatic model selection.

KW - Analytics

KW - Feature selection

KW - Support vector regression

KW - Time series analysis

KW - Electric load forecasting

U2 - 10.1016/j.asoc.2019.105616

DO - 10.1016/j.asoc.2019.105616

M3 - Journal article

VL - 83

JO - Applied Soft Computing Journal

JF - Applied Soft Computing Journal

SN - 1568-4946

M1 - 105616

ER -

Research

Links

Text available via DOI:

Keywords