Forecasting tourists’ characteristics by a genetic algorithm with a transition matrix

The number of tourists that visit a mature tourism destination becomes stable. If recommendations that could adjust the supply structure towards demand needs in this scenario are established as administration goals, then forecasting the total number of tourists is important, but of particular interest is the internal composition of the demand in terms of specific characteristics of the future groups. This view type of forecasts can be achieved by a genetic algorithm (GA). In this paper, a GA with a transition probability matrix is developed and, as an illustration of the methodology, one run of such an algorithm has achieved better forecasting performance than a simple GA.