FORECASTING DYNAMIC TOURISM DEMAND USING ARTIFICIAL NEURAL NETWORKS

Planning tourism development means preparing the destination for coping with uncertainties as is sensitive to many changes. This study tested two types of artificial neural networks in modeling international tourist arrivals recorded Ohrid (North Macedonia) during 2010–2019. It argues that MultiLayer Perceptron (MLP) network more accurate than Nonlinear AutoRegressive eXogenous (NARX) model when forecasting demand. The research reveals bigger number neurons may not necessarily lead further perfor- mance improvement model. MLP its better performance series unexpected challenges highly recommended dynamic demand