Home > Research > Publications & Outputs > A neural network approach for the Theta model

Research

Associated organisational unit

Links

Keywords

View graph of relations

A neural network approach for the Theta model

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

Published

Standard

A neural network approach for the Theta model. / Constantinidou, Christina; Nikolopoulos, Konstantinos; Bougioukos, Nikolaos et al.
In: Lecture Notes in Information Technology, Vol. 25, 2012, p. 116-120.

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

Harvard

Constantinidou, C, Nikolopoulos, K, Bougioukos, N, Tsiafa, E, Petropoulos, F & Assimakopoulos, V 2012, 'A neural network approach for the Theta model', Lecture Notes in Information Technology, vol. 25, pp. 116-120. <http://www.ier-institute.org/2070-1918/lnit25/>

APA

Constantinidou, C., Nikolopoulos, K., Bougioukos, N., Tsiafa, E., Petropoulos, F., & Assimakopoulos, V. (2012). A neural network approach for the Theta model. Lecture Notes in Information Technology, 25, 116-120. http://www.ier-institute.org/2070-1918/lnit25/

Vancouver

Constantinidou C, Nikolopoulos K, Bougioukos N, Tsiafa E, Petropoulos F, Assimakopoulos V. A neural network approach for the Theta model. Lecture Notes in Information Technology. 2012;25:116-120.

Author

Constantinidou, Christina ; Nikolopoulos, Konstantinos ; Bougioukos, Nikolaos et al. / A neural network approach for the Theta model. In: Lecture Notes in Information Technology. 2012 ; Vol. 25. pp. 116-120.

Bibtex

@article{48038a39f573434ea4ec320c1f2179d8,
title = "A neural network approach for the Theta model",
abstract = "Classic Theta model decomposes the original data series into two separate lines, which are extrapolated separately and the forecasts are combined with equal weights. The current study explores a neural network approach to Theta model, in terms of optimizing the combination weights of the two components in the final forecast. The performance of the proposed method (Theta AI) is compared against the original method for the two subsets of the NN3 forecasting competition, which primary objective was the evaluation of methods using neural networks or artificial intelligence for time series forecasting. Results indicate that the new approach is very promising towards the generalization of the Theta model.",
keywords = "Time Series Forecasting, Theta Model, Neural Networks, Artificial Intelligence, Forecasting Competitions, Theta AI ",
author = "Christina Constantinidou and Konstantinos Nikolopoulos and Nikolaos Bougioukos and Eirini Tsiafa and Fotios Petropoulos and Vassilios Assimakopoulos",
year = "2012",
language = "English",
volume = "25",
pages = "116--120",
journal = "Lecture Notes in Information Technology",
issn = "2070-1918",

}

RIS

TY - JOUR

T1 - A neural network approach for the Theta model

AU - Constantinidou, Christina

AU - Nikolopoulos, Konstantinos

AU - Bougioukos, Nikolaos

AU - Tsiafa, Eirini

AU - Petropoulos, Fotios

AU - Assimakopoulos, Vassilios

PY - 2012

Y1 - 2012

N2 - Classic Theta model decomposes the original data series into two separate lines, which are extrapolated separately and the forecasts are combined with equal weights. The current study explores a neural network approach to Theta model, in terms of optimizing the combination weights of the two components in the final forecast. The performance of the proposed method (Theta AI) is compared against the original method for the two subsets of the NN3 forecasting competition, which primary objective was the evaluation of methods using neural networks or artificial intelligence for time series forecasting. Results indicate that the new approach is very promising towards the generalization of the Theta model.

AB - Classic Theta model decomposes the original data series into two separate lines, which are extrapolated separately and the forecasts are combined with equal weights. The current study explores a neural network approach to Theta model, in terms of optimizing the combination weights of the two components in the final forecast. The performance of the proposed method (Theta AI) is compared against the original method for the two subsets of the NN3 forecasting competition, which primary objective was the evaluation of methods using neural networks or artificial intelligence for time series forecasting. Results indicate that the new approach is very promising towards the generalization of the Theta model.

KW - Time Series Forecasting

KW - Theta Model

KW - Neural Networks

KW - Artificial Intelligence

KW - Forecasting Competitions

KW - Theta AI

M3 - Journal article

VL - 25

SP - 116

EP - 120

JO - Lecture Notes in Information Technology

JF - Lecture Notes in Information Technology

SN - 2070-1918

ER -