Modelling Air Passenger Traffic Flow in Murtala Muhammad International Airport Lagos, Nigeria: A Time Series Approach

Abstract

In this study, Artificial Neural Network (ANN), Seasonal Auto-Regressive Integrated Moving Average (SARIMA) and Holt-Winters Exponential Smoothing (HWES) models are used to model air passenger traffic flow in Murtala Muhammad International Airport (MMIA), Lagos Nigeria. The performances of these proposed models are compared for in-sample and out-of-sample performance by employing static forecast procedure over January 2014 to December 2015 forecast horizon. The best models from the SARIMA, ANN and HWES were selected by employing some performance metrics comprising, Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) and residual diagnostics. The selected models forecasting performances were compared using the statistical loss functions, Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) for the measurement of forecast accuracy. Results show that ANN outperforms the other models in the domestic sector, while the HWES had the best performance in the international sector even though it was outperformed by SARIMA in the domestic sector. ANN yielded the best in-sample performance for domestic and international air passenger traffic. It was concluded that the ANN, which represents a class of non-linear time series model is very efficient in mimicking time series pattern and giving good forecast.