抄録
Though service sector in Japan has developed, number of traditional inns (ryokan) has declined over the last 30 years. One of the reasons for this decline is that many ryokan managers tend to rely more on experience and intuition than on data analysis. Moreover, few studies have focused on hotel inventory control. Therefore, considering that safety stock levels and inventory costs have become crucial factors in the hotel industry, this study proposes methods for determining an appropriate level of safety stock using demand forecasting that considers the peculiarities of the hotel business. It applies multiple regression analysis and neural networks to forecast beer demand based on sales data from company A, a traditional Japanese inn. This study also compares the proposed safety stock level with that of company A. Furthermore, it provides an empirical analysis of inventory management in the hotel business. The results not only demonstrate that the reorder criteria determined by the multiple regression analysis and ANN are superior to the method currently used by company A but also verify the efficacy of the proposed methods.
