Predicting Passenger Flow at Charles De Gaulle Airport Security Checkpoints

Abstract

Airport security checkpoints are critical areas in airport operations. Airports have to manage an important passenger flow at these checkpoints for security reason while maintaining service quality. The cost and quality of such an activity depend on the human resource management for these security operations. An appropriate human resource management can be obtained using an estimation of the passenger flow. This paper investigates the prediction at a strategic level of the passenger flows at Paris Charles De Gaulle airport security checkpoints using machine learning techniques such as Long Short-Term Memory neural networks. The derived models are compared to the current prediction model using three different mathematical metrics. In addition, operational metrics are also designed to further analyze the performance of the obtained models.