A Study on Adaptive Forecasting of Bakery Sales

• Main Authors: Chih-Hui Huang, 黃智慧
• Other Authors: Mei-Ting Tsai
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/n4y56m

碩士 === 國立中興大學 === 高階經理人碩士在職專班 === 106 === Sales forecasting is the basis for all corporate strategy and decision analysis. If the sales forecast is not accurate, some problems may arise. For example, the production process is not smooth, inventory is not properly controlled, transportation costs and obsolete inventory may increase, and the operating costs of the company may increase. Therefore, the purpose of this study is to establish a sales forecasting model in the baking industry. This study conducts an empirical example in the baking industry and collects sales data over the past 40 months. At the same time, it measures the impact of trends or seasonal factors, adopts an exponential smoothing method of trend or seasonal correction, and establishes appropriate sales forecasting models. It also calculates the prediction error and mean absolute percentage error (MAPE), studies the performance index of the most tolerable error rate, in order to verify whether the benefit of the forecast model is better than the current situation and solves the enterprise’s problem. From the results of the study, the use of adaptive forecasting methods to predict the number of sales can increase the accuracy of forecasting and significantly reduce the shortage of goods compared to personal experience judgments. At the same time, the emergency replenishment situation can also be significantly reduced. Companies need to properly use predictive models and often monitor and record errors. Periodically, companies analyze the causes of errors and adjust prediction models appropriately. If companies make good use of computer systems for large-scale calculations, then the predictions will help optimize the operating efficiency.