Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Cited by Google
- Similars in SciELO
- Similars in Google
Share
Estudios Gerenciales
Print version ISSN 0123-5923
Abstract
VELASQUEZ, JUAN DAVID; FRANCO, CARLOS JAIME and GARCIA, HERNÁN ALONSO. UN MODELO NO LINEAL PARA LA PREDICCIÓN DE LA DEMANDA MENSUAL DE ELECTRICIDAD EN COLOMBIA*. estud.gerenc. [online]. 2009, vol.25, n.112, pp.37-54. ISSN 0123-5923.
A non-linear model for forecasting the monthly demand for electricity in Colombia This article provides a comparison of the performance of an ARIMA model, a multilayer perceptron, and an autoregressive neural network for forecasting the monthly demand for electricity in Colombia for the following month. The available data were divided into two different sets, i.e. one set for estimating the model parameters, and the other for evaluating the forecast ability outside the range of the sample calibration data. The results show that the autoregressive neural network is able to forecast the demand more accurately than the other two models when the total available data are considered.
Keywords : Demand; forecast; neural networks; ARIMA.