Development of a Predictive Analytics Solution for Forecasting Rice Demand in Quezon Province Using Multiple Regression Algorithm

The study focuses on the development of predictive analytic solution for forecasting rice demand in Quezon. Inaccurate rice demand forecast could lead to incorrect volume and ill-timing of rice imports which could result in either a waste of resources for the government or a burden to consumers especially the food establishments and households which consider rice as one of the primary commodity. The data used in the study came from the Office of the Provincial Agriculturist and Philippine Statistics Authority that served as predictors used in the system. Multiple regression algorithm is used to forecast the demand of rice in Quezon province together with the predictors used like year, food wastage, net production of rice and demand per capita. CRISP-DM was used as software development methodology in fulfilling the requirement method and business objectives of the study. The forecast results of the system was compared to the results of the data analysis feature of Microsoft Excel and also the training and testing of datasets which show that the forecasted demand is close to the actual demand of rice.