A Decision Support System for Hotel Facilities Inventory Management

A major goal of a tourism supply chain is a profitable collaboration between actors involved. Small hotel facilities tend to order small amounts of each good. The unit cost is generally unfavorable compared to that of large hotel facilities. To overcome this disadvantage, small facilities can collaborate by placing aggregate orders to a single vendor. Consequently, the increased quantity ordered can afford a unit cost reduction. This paper investigates the effectiveness of a set of novel demand forecast techniques for supporting this order aggregation. We describe four different algorithms: they all use the orders' history; in addition, two of them forecast the numbers of guests. The performed tests use large amount of anonymous real-world data and show that the algorithms that also use the numbers of guests performs better than those based only on the orders' history.