Sale forecast for basic commodities based on artificial neural networks prediction

Abstract

The objective of this paper is to carry out the comparison and selection of a method to forecast sales of basic food products efficiently. The source of data comes from a set of popular markets in the main departments of Colombia. The methods and methodologies used are: Hold Method, Winters, the Box Jenkins methodology (ARIMA) and an Artificial Neural Network. The results show that the artificial neural network obtained a better performance achieving the lowest mean square error.