Forecasting air travel demand of Kuwait: A comparison study by using regression vs. artificial intelligence

Air traffic demand is an important factor in the planning, design and construction of airport facilities. Although population, national income, import, export and the numerical strength of the labor force have rapidly increased in the last ten years in Kuwait; an analysis of the current impact of these factors on air traffic demand is lacking in literature. The main objectives of this study were to (a) determine the main factors that affect air traffic demand in Kuwait; (b) calibrate a regression model regressing air demand and the causal factors, (c) to apply artificial intelligence software (NeuroShell) as a technique for forecasting air travel demand, and (d) compare the regression and artificial intelligence model results, and select the best models to forecast air travel demand for Kuwait. Findings have indicated that national income and labor force were the main factors affecting air travel demand in Kuwait. The developed models all predicted the demand for air travel with high accuracy. The (NeuroShell) models were selected as the choice of forecasting model for Kuwait due to their better goodness of fit over other regression models.