Paper Title
Domestic And International Air Transportation Passenger Demand Prediction With Flower Pollination Algorithm: The Case Of Turkey

Abstract
After 1980, Turkey has shown significant improvements in air transport and it has been a notable increase in demand for air transport last 16 years.The total number of passengers (domestic line + international line) is increased approximately 6 times.The increase in economic levels in Turkey supports the continuation of the upward trend in air transport demand. In this study, passenger demand forecasting models for domestic and international flights of the air transport in Turkey were developed and future passenger demands were predicted according to created scenario. Depending on 4 independent parameters, 3 different models were developed in quadratic, exponential and force forms for domestic and international flights separately. As input parameters of models, the total number of seats, Gross Domestic Product (GDP) per capita($), annual jet fuel consumption (barrel/year) and jet fuel price ($ / barrel) between 2002 and 2017 were used and the models were optimized using Flower Pollination Algorithm(FPA).The accuracy of the models was presented based on performance criteria by comparing to the observed values.As the performance criteria, the coefficient of determination (R2) and the mean absolute percentage errors (MAPE)were used. Although the quadratic model was the best result, the force model was preferred for future prediction in terms of the practicality of the model. According to the created scenario, the passenger demands up to 2030 were forecasted with power model.It is predicted that the number of passengers who will be carried by air transport in 2030 will increase 4 times for the domestic and approximately 2.5 times for the international flights when it is compared to the year 2017. Keywords - Passenger Demand Prediction Models, Flower Pollination Algorithm, Turkey