Utilização de redes neurais recorrentes para predição de controle de estoque

Abstract

he present work aims at studying the Recurrent Neural network (RNN) as a method for prediction of the amount of stock needed for the given period of maize deseeded. Being that the corn product marketed for a specific seller of agricultural products. Before that, it was made a study of the historical series of sales of the product in question and so confronted past periods for the prediction in future periods.