Procedimiento para el pronóstico de la demanda mediante redes neuronales artificiales

In the planning process, forecasts have an inescapable role, being the first step of this process and on which the results to be obtained will depend. This article proposes a procedure to perform the forecast of the demand through Artificial Neural Networks. In order to carry out the practical validation, the proposed procedure was applied to the Holguin Marketing and Distribution Company (EMCOMED) and the following goods were selected: Captopril 25 mg 10 tab, Dipirone 300 mg 10 tab and Heberprot-P 25 and 75. The results prove that the proposed procedure is reliable enough to become an easily applied resource.