Design and Implementation of Inventory Forecasting System using Double Exponential Smoothing Method

摘要

Recently, information systems have an essential role in companies, especially in companies that engaged in the production of goods and services. Prediction or forecasting is a mandatory supporting component in planning activities in making business predictions for maximum profit. To tackle this issue, this paper proposes to design and implement an inventory forecasting system. The proposed system utilizes the Double Exponential Smoothing forecasting method, which is a form of quantitative inventory control based on the historical data (time series) so that the movement of data from the past can be analyzed. In the end, the proposed system may computationally draw the movement trends in the future. The forecasting itself employs two parameters, alpha $(lpha)$ and beta ($eta$). In this study, two values are observed to obtain the best form of optimal forecasting results for the future with the minimum error rate. In this study, forecasting experiments are conducted by using one of the examples of the stock of goods with a range of five periods based on months by using the parameter values alpha $(lpha)=0.5$ and beta $(eta)=0.5$. After conducting the number of experiments, it is shown from testing that the obtained MAPE (Mean Absolute Percentage Error) showed decent results, which is around 33.18%.