Robust electricity consumption modeling of Turkey using Singular Value Decomposition

Abstract

Multivariable regression method is used to model Turkey's electricity consumption through a nonlinear relationship. Electricity consumption is modeled as a function of four demographic and economic indicators such as, population, gross domestic product per capita, imports and exports. The second order model includes 15 coefficients for bias, first degree terms and second degree terms. Data preprocessing is applied to transform all variables to have zero mean and percent relative variance. Singular Value Decomposition is applied to reduce the dimensionality of the problem and to provide robustness to the estimations. Variance and covariance information in the data set is used to determine the number of important dimensions in the data. Electricity consumption of Turkey is modeled using annual data from 1970 to 2011. The results show that electricity consumption can be robustly modeled using Singular Value Decomposition. © 2013 Elsevier Ltd. All rights reserved.