FORECASTING IMPORT DATA USING NON-LINEAR MODEL

When Malaysia government implemented GST in 2015, it raised the import of Malaysia as GST minimized the production of local goods and drove up costs. But, too many imports in relation to exports can distort a nation?s balance of trade and devalue its currency. In order to lower the possibility of the increasing imports, it is important to determine the future import value in advance. If the future import is forecasted, then action can be taken to reduce the consequence effects of the high imports. In this study, we predict the future import value using five empirical models of least square method: Linear model, Logarithmic model, Power model, Exponential model and Polynomial model. The method has shown that Quadratic model is the best fitted model for Malaysia import data.