Studying Long Range Dependence of Indigo Air Traffic Using ARFIMA

Abstract

  In the present study, the time series models ARIMA and ARFIMA or FARIMA models have been fitted to Air India domestic air passengers, which considered as self similarity and Long Range Dependence (LRD). In such case ARFIMA model is expected to be superior to ARIMA. We fitted ARIMA and ARFIMA models to air traffic data and compared. Then the best model has been identified using RMSE, MAE and MAPE values. This model can be useful to analyze the air traffic flow and revise the services of Indigo. The analysis was carried out using time series data on number of passengers travelling by Air India domestic flights during January 2012 to December 2018