An Application of Forecasting Models for the Supply and Demand Management of Cassava Products

Abstract

The objectives of this research are to generate models that can effectively forecast the supply and demand of four cassava products. The appropriate forecasting models for cassava production volume is Back Propagation Neural Network (BPN) 4-14-1, cassava starch is BPN 7-12-1, cassava chip is BPN 7-14-1, cassava pellets is Multiple Linear Regression (MLP), and sago is BPN 7-13-1. Then, Linear Programming is used to calculate the optimization of cassava products to obtain the maximum profit and for cassava plant areas to obtain the maximum yield per area. The benefits of this research can support management planning for farmers and manufacturers.