Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling

Wahyu Wibowo; Sarirazty Dwijantari; Alia Hartati

doi:10.1007/978-981-10-7242-0_11

Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling

Wahyu Wibowo^*
, Sarirazty Dwijantari
, Alia Hartati

^*Corresponding author for this work

Laboratory of Quality and Productivity Engineering

Institut Teknologi Sepuluh Nopember

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

The aims of this paper are to develop a linear and nonlinear model in time series to forecast electricity consumption of the lowest household category in East Java, Indonesia. The installed capacity in the lowest household customer category has various power, i.e. 450 VA, 900 VA, 1300 VA, and 2200 VA. ARIMA models are family of linear model for time series analysis and forecasting for both stationary and non-stationary, seasonal and non-seasonal time series data. A nonlinear time series model is proposed by hybrid ARIMA-ANN, a Radial Basis Function using orthogonal least squares. The criteria used to choose the best forecasting model are the Mean Absolute Percentage Error and the Root Mean Square Error. The ARIMA best model are ARIMA ([1, 2], 1, 0) (0, 1, 0)¹², ARIMA (0, 1, 1) (0, 1, 0)¹², ARIMA (0, 1, 1) (0, 1, 0)¹², ARIMA (1, 0, 0) (0, 1, 0)¹² respectively. The ANN architecture optimum are ANN (2, 12, 1), ANN (1, 12, 1), ANN (1, 12, 1), and ANN (1, 12, 1). The best models are ARIMA ([1, 2], 1, 0) (0, 1, 0)¹², ARIMA (0, 1, 1) (0, 1, 0)¹², ANN (1, 12, 1), and ANN (1, 12, 1) in each category respectively. Hence, the result shows that a complex model is not always better than a simpler model. Additionally, a better hybrid ANN model is relied on the choice of a weighted input constant of RBF.

Original language	English
Title of host publication	Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings
Editors	Azlinah Mohamed, Bee Wah Yap, Michael W. Berry
Publisher	Springer Verlag
Pages	126-139
Number of pages	14
ISBN (Print)	9789811072413
DOIs	https://doi.org/10.1007/978-981-10-7242-0_11
Publication status	Published - 2017
Event	3rd International Conference on Soft Computing in Data Science, SCDS 2017 - Yogyakarta, Indonesia Duration: 27 Nov 2017 → 28 Nov 2017

Publication series

Name	Communications in Computer and Information Science
Volume	788
ISSN (Print)	1865-0929

Conference

Conference	3rd International Conference on Soft Computing in Data Science, SCDS 2017
Country/Territory	Indonesia
City	Yogyakarta
Period	27/11/17 → 28/11/17

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

ARIMA
Electricity consumption
Forecasting
Hybrid ARIMA-ANN

Access to Document

10.1007/978-981-10-7242-0_11

Cite this

Wibowo, W., Dwijantari, S., & Hartati, A. (2017). Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling. In A. Mohamed, B. W. Yap, & M. W. Berry (Eds.), Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings (pp. 126-139). (Communications in Computer and Information Science; Vol. 788). Springer Verlag. https://doi.org/10.1007/978-981-10-7242-0_11

Wibowo, Wahyu ; Dwijantari, Sarirazty ; Hartati, Alia. / Time series machine learning : Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling. Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings. editor / Azlinah Mohamed ; Bee Wah Yap ; Michael W. Berry. Springer Verlag, 2017. pp. 126-139 (Communications in Computer and Information Science).

@inproceedings{7c3f9e4a627a41429dc59a12c653558b,

title = "Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling",

abstract = "The aims of this paper are to develop a linear and nonlinear model in time series to forecast electricity consumption of the lowest household category in East Java, Indonesia. The installed capacity in the lowest household customer category has various power, i.e. 450 VA, 900 VA, 1300 VA, and 2200 VA. ARIMA models are family of linear model for time series analysis and forecasting for both stationary and non-stationary, seasonal and non-seasonal time series data. A nonlinear time series model is proposed by hybrid ARIMA-ANN, a Radial Basis Function using orthogonal least squares. The criteria used to choose the best forecasting model are the Mean Absolute Percentage Error and the Root Mean Square Error. The ARIMA best model are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (1, 0, 0) (0, 1, 0)12 respectively. The ANN architecture optimum are ANN (2, 12, 1), ANN (1, 12, 1), ANN (1, 12, 1), and ANN (1, 12, 1). The best models are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ANN (1, 12, 1), and ANN (1, 12, 1) in each category respectively. Hence, the result shows that a complex model is not always better than a simpler model. Additionally, a better hybrid ANN model is relied on the choice of a weighted input constant of RBF.",

keywords = "ARIMA, Electricity consumption, Forecasting, Hybrid ARIMA-ANN",

author = "Wahyu Wibowo and Sarirazty Dwijantari and Alia Hartati",

note = "Publisher Copyright: {\textcopyright} Springer Nature Singapore Pte Ltd. 2017.; 3rd International Conference on Soft Computing in Data Science, SCDS 2017 ; Conference date: 27-11-2017 Through 28-11-2017",

year = "2017",

doi = "10.1007/978-981-10-7242-0\_11",

language = "English",

isbn = "9789811072413",

series = "Communications in Computer and Information Science",

publisher = "Springer Verlag",

pages = "126--139",

editor = "Azlinah Mohamed and Yap, \{Bee Wah\} and Berry, \{Michael W.\}",

booktitle = "Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings",

address = "Germany",

}

Wibowo, W, Dwijantari, S & Hartati, A 2017, Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling. in A Mohamed, BW Yap & MW Berry (eds), Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings. Communications in Computer and Information Science, vol. 788, Springer Verlag, pp. 126-139, 3rd International Conference on Soft Computing in Data Science, SCDS 2017, Yogyakarta, Indonesia, 27/11/17. https://doi.org/10.1007/978-981-10-7242-0_11

Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling. / Wibowo, Wahyu; Dwijantari, Sarirazty; Hartati, Alia.
Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings. ed. / Azlinah Mohamed; Bee Wah Yap; Michael W. Berry. Springer Verlag, 2017. p. 126-139 (Communications in Computer and Information Science; Vol. 788).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Time series machine learning

T2 - 3rd International Conference on Soft Computing in Data Science, SCDS 2017

AU - Wibowo, Wahyu

AU - Dwijantari, Sarirazty

AU - Hartati, Alia

N1 - Publisher Copyright: © Springer Nature Singapore Pte Ltd. 2017.

PY - 2017

Y1 - 2017

N2 - The aims of this paper are to develop a linear and nonlinear model in time series to forecast electricity consumption of the lowest household category in East Java, Indonesia. The installed capacity in the lowest household customer category has various power, i.e. 450 VA, 900 VA, 1300 VA, and 2200 VA. ARIMA models are family of linear model for time series analysis and forecasting for both stationary and non-stationary, seasonal and non-seasonal time series data. A nonlinear time series model is proposed by hybrid ARIMA-ANN, a Radial Basis Function using orthogonal least squares. The criteria used to choose the best forecasting model are the Mean Absolute Percentage Error and the Root Mean Square Error. The ARIMA best model are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (1, 0, 0) (0, 1, 0)12 respectively. The ANN architecture optimum are ANN (2, 12, 1), ANN (1, 12, 1), ANN (1, 12, 1), and ANN (1, 12, 1). The best models are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ANN (1, 12, 1), and ANN (1, 12, 1) in each category respectively. Hence, the result shows that a complex model is not always better than a simpler model. Additionally, a better hybrid ANN model is relied on the choice of a weighted input constant of RBF.

AB - The aims of this paper are to develop a linear and nonlinear model in time series to forecast electricity consumption of the lowest household category in East Java, Indonesia. The installed capacity in the lowest household customer category has various power, i.e. 450 VA, 900 VA, 1300 VA, and 2200 VA. ARIMA models are family of linear model for time series analysis and forecasting for both stationary and non-stationary, seasonal and non-seasonal time series data. A nonlinear time series model is proposed by hybrid ARIMA-ANN, a Radial Basis Function using orthogonal least squares. The criteria used to choose the best forecasting model are the Mean Absolute Percentage Error and the Root Mean Square Error. The ARIMA best model are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ARIMA (1, 0, 0) (0, 1, 0)12 respectively. The ANN architecture optimum are ANN (2, 12, 1), ANN (1, 12, 1), ANN (1, 12, 1), and ANN (1, 12, 1). The best models are ARIMA ([1, 2], 1, 0) (0, 1, 0)12, ARIMA (0, 1, 1) (0, 1, 0)12, ANN (1, 12, 1), and ANN (1, 12, 1) in each category respectively. Hence, the result shows that a complex model is not always better than a simpler model. Additionally, a better hybrid ANN model is relied on the choice of a weighted input constant of RBF.

KW - ARIMA

KW - Electricity consumption

KW - Forecasting

KW - Hybrid ARIMA-ANN

UR - https://www.scopus.com/pages/publications/85036460914

U2 - 10.1007/978-981-10-7242-0_11

DO - 10.1007/978-981-10-7242-0_11

M3 - Conference contribution

AN - SCOPUS:85036460914

SN - 9789811072413

T3 - Communications in Computer and Information Science

SP - 126

EP - 139

BT - Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings

A2 - Mohamed, Azlinah

A2 - Yap, Bee Wah

A2 - Berry, Michael W.

PB - Springer Verlag

Y2 - 27 November 2017 through 28 November 2017

ER -

Wibowo W, Dwijantari S, Hartati A. Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling. In Mohamed A, Yap BW, Berry MW, editors, Soft Computing in Data Science - 3rd International Conference, SCDS 2017, Proceedings. Springer Verlag. 2017. p. 126-139. (Communications in Computer and Information Science). doi: 10.1007/978-981-10-7242-0_11

Time series machine learning: Implementing ARIMA and hybrid ARIMA-ANN for electricity forecasting modeling

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this