Auto parts demand forecasting based on nonnegative variable weight combination model in auto aftermarket

Abstract

Accurate demand forecasting for auto parts can improve the performance of the whole auto supply chain and is very important for the management improvement for the companies in auto aftermarket who mainly forecast demands by experience. It has both economic significance and social means for the auto industry considering the important role of auto aftermarket in the whole auto industry. After exploring the complicated characteristics of the auto parts and also the strengths of some forecasting methods, ARIMA, multiple regression and Support Vector Regression are selected finally to develop a nonnegative variable weight combination model to forecast the demand of auto parts for the auto aftermarket in China. The following case study proves that this model has higher accuracy and more stability.