Forecasting air passenger travel: A case study of Norwegian aviation industry

Abstract

Accurate forecasting of airline passenger traffic is important for facilitating the effective management and planning of aviation resources. In this study, we explore the air passenger traffic in the Norwegian aviation industry by collecting the passenger flow data and the corresponding measurements of the weather conditions affecting the flow from the different airports in Norway. We then proposed nonlinear autoregressive with exogenous input (NARX) forecasting models to predict air passenger traffic in advance. The NARX models account for the nonlinearity and nonstationarity in the passenger flow and allow the accurate forecasting of air passenger traffic. We perform experiments to demonstrate the effectiveness of two variants of the NARX model and compare their performances against long short-term memory (LSTM), a deep learning method. We show that the proposed NARX model achieves the best prediction accuracy compared to LSTM, which is considered as a state-of-the-art approach for fitting sequential data.