Forecasting Model for Product Sales in Electric Appliance Category: A Case Study of Topvalue Corporate Company Limited

The objective of this project is to find out the appropriate forecasting model for two product sales, induction cooker and coffee maker, which are the top 10 bestselling products in electric appliance category of Topvalue Corporate Company Limited. The data for each set is weekly product sale during January 2019 to December 2020 of 104 values. The data are split into 2 sets, the first 92 values for creating the forecasting model and the last 12 values for checking the accuracy of the forecasting model using mean absolute percentage error (MAPE). Compare three forecasting models which are Smoothing Method, Box-Jenkins Method, and Demand forecasting Method. The results show that the appropriate model for forecasting product sale of induction cooker is Box-Jenkins Method, ARIMA(1,1,1) and the appropriate model for forecasting product sale of coffee maker is  Croston’s Method.