Apply Machine Learning In Sales Prediction And Aggregate Planning: A Case Study Of The Automobile Industry

Abstract

Nowadays, sales of the automobile industry the in Vietnam market have grown exponentially over the years. Therefore, the demand for car products is also increasing over time, this requires a timely response from suppliers. This paper is aimed at solving a problem domain of the automobile industry, which plays a key vital to reduce costs due to the high cost of accessories and the development of this industry in these years. To determine which forecasting method is most appropriate for the specific demand, SES is used in contrast to ARIMA. However, achieving good sales would need improvements in demand forecasting and in the production planning process. A multi-objective mathematical model is built in this paper with the goal is maximize the profit, minimize the other factors such as rate of raw material, backorder quantity and total overtime are in order to adapt the fluctuation forecasted demand in a period of 12 months from the above two forecasting techniques. By comparing with hisotical data, decision makers can give better recommendations for organizations regarding the selection of the best production forecasting model in order to satisfy client demand.