Cluster Analysis of World's Airports on the Basis of Number of Passengers Handled (Case Study Examinning the Impact of Significant Events)

Abstract

Nowadays, the air transportation is one of key means of transport. Unfortunately, there are many factors influencing its popularity and intensity of its use. There are many studies investigating these factors. The paper investigates the possibility of classifying the world's airports in terms of the trend in the number of handled passengers as it is one of the main economic indicators for airports. For this classification we chose cluster analysis. The paper focuses on an important aspect of the process, which chooses the appropriate number of clusters. It turned out that in terms of interpretation of the results, it may not always be the most efficient to set this number at the theoretically best and recommended value. As a result of our analysis, several groups of airports with similar trend of post-event reactions are found. Therefore, this may bring better understanding of the intensity and the range of the impact of particular events on air transportation.

Keywords

• Cluster analysis
• number of clusters
• occupancy of airports
• Bayesian Information Criterion
• Akaike Information Criterion
• silhouette coefficient