Modeling and Forecasting of Tourist Arrivals in Crete Using Statistical Models and Models of Computational Intelligence: A Comparative Study

The Forecast Of Tourist Arrivals

Because of the aforementioned significant contribution of tourism to the local economy and the nature of the tourist industry, the design of the tourism policy is crucial to the whole process of tourism development. It is widely accepted that any business design regarding tourism requires some form of decision-making, based on future forecasting (Palmer et al., 2006, Chandra & Menezes, 2001). Thus, the reliable and the accurate forecast of the tourism demand is essential in making decisions for effective and efficient planning of tourism policy. Moreover, the accurate forecasting of tourism demand is necessary for the successful design of the supply of goods, services and the infrastructures in tourism and in other service industries. Optimal forecasts of tourist arrivals in a country or in a particular geographical area, can effectively help business managers and investors to make operational, tactical and strategic decisions. These may include decisions on scheduling and on personnel management, on preparing tourist booklets and hotel investments etc. Furthermore, government councils need accurate tourist forecasting in order to design the required tourist infrastructure, such as new hotels in touristic places, as well as the creation of appropriate transport and communications. It is also necessary to predict the consumption of goods and services by tourists, in order to assess the relative contribution of tourism to production, income and employment in tourist destinations (Palmer et al., 2006).

The forecast configuration based generally on two basic conditions. The first one regards the quantitative method that is used and the second is how the economic variable behaves. The better the way the variables are created, the better predictions are expected to be formed. However, for this to be the case, the existing structure of the behavior of the variable's value should remain constant or almost stable in the near future.