Two Forecasting Models for Predicting the Market Sales of Fresh Milk in Taiwan

• Main Authors: Jeffrey Pan, 潘家銳
• Other Authors: 游張松
• Format: Others
• Language: en_US
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/43368181668855684193

碩士 === 國立臺灣大學 === 國際企業管理組 === 99 === ABSTRACT Accurate demand forecasts result in high levels of customer service and efficient operations, while inaccurate forecasts tend to lead to poor levels of customer satisfaction and higher cost operations. In many instances, the first step we can do to improve the efficiency and effectiveness of a company is to improve the quality of the market sales forecasts. Estimation of seasonal demand prior to an active demand season is essential. The demand of seasonal products frequently changes in the marketplace. As soon as the main selling season passes, the excessive inventories of the product are devalued greatly. Furthermore, if the product supplies are relatively short, a direct loss in sales occurs. Therefore, demand planning is considered the fundamental process of a business plan, which provides a continuous link to manage the inventory position and the product demand. It will often be the case that the market sales possess a pattern that includes both trend and seasonality. The focus of this research is to predict the future market sales of a seasonal product such as fresh milk using Winter’s multiplicative trend seasonal model and the Decomposition method. This research examines and compares the results of both the Winter’s multiplicative trend seasonal model and the decomposition method to develop the better forecast for the seasonal demand. Results indicate that both models are well fitting ones. This study improves the quality of sales forecasts and facilitates in determining the better forecast. Results also imply, theoretically, that the decomposition method has significant advantages in terms of accuracy.