國際觀光業是世界最大及成長最快速的產業，由於觀光產業的產品與服務具有無法保存的特性，因此如果能準確預測國際觀光客人數便可對觀光產品、服務的提供及基本的設施做成功的規劃以增加經濟利益。因此本研究提出一個新的預測工具，即是使用以Widrow-Hoff學習演算法為基礎的可適性模糊網路來建立一個預測的模式，並據以預測來台的日本與美國觀光客人數，其預測結果可以達到非常不錯的準確度，因此可證實此研究所建立之模式是一個具有良好預測能力的模式，並可作為觀光產業決策者與管理者在觀光產品及設施規劃上的重要參考。
International tourism has become one of the largest and most rapidly growing industries in the world. Since there exists the perishable nature of the product and service in the tourism industry, it is crucial to have an accurate forecast of its international visitors and tourism receipts in order to choose an appropriate strategy for its economic benefits. In this paper, a new approach is proposed and that is a fully connected adaptive fuzzy network (AFN) based on Widrow-Hoff learning algorithm to model and forecast the tourist arrivals for the travel of international visitors to Taiwan. And the difference between the expected and the forecast output values falls into a very acceptable range of discrepancies, which means that using the adaptive fuzzy network has reached the required level of accuracy. The result is in good accord with the monitored data and allows its use as the forecasting model to help policy makers and managers of tourism industry to develop planning for various tourism activities.
