Most of the time the nature of the traffic is bursty. This traffic congestion may be studied mathematically as LRD (long-range dependence) or self-similarity. The subject of the analysis is Indigo with the number of passengers travelled by Domestic flights being the focus. The analysis utilized time series data on the number of passengers traveling on Air India domestic flights from January 2012 to December 2018. The data has been taken from the Directorate General of Civil Aviation (DGCA) website. SPSS and STATA software were used for this analysis. This study fitted the time series models ARIMA and ARFIMA (or FARIMA) to flight Indian domestic flight passenger data, which exhibit self-similarity and Long Range Dependence (LRD). In this scenario, the ARFIMA model is anticipated to outperform the ARIMA model. ARIMA and ARFIMA models were applied to air traffic data and a comparison was conducted. The optimal model has been determined utilizing RMSE, MAE, and MAPE metrics. This model can effectively assess air traffic flow and optimize Indigo's services. The results show that RMSE, MAE and MAPE are lower for the ARFIMA model than ARIMA. Also, the R-squared value is greater for the ARFIMA model than ARIMA, which shows ARFIMA model is superior to ARIMA for this data. In conclusion, predictions regarding the future air traffic of Indigo domestic carriers may be made with the help of the ARFIMA model as it is beneficial for Indigo to adjust their services with the help of this study.
Author (s) Details
Manohar Dingari
Anurag University, Hyderabad, India.
V. Sumalatha
Veeranari Chakari Ilamma Women’s University, Koti, Hyderabad, India.
S. Hariprasad
Anurag University, Hyderabad, India.
S. Lavanya
Visvesvaraya College of Engineering and Technology, Hyderabad, India.
Please see the book here:- https://doi.org/10.9734/bpi/stda/v9/5117
