Forecasting of Air Passengers using ARIMA Modeling
Sakshi Sharma1, A.Jackulin Mahariba2
1Sakshi Sharma, SRM Institute of science and Technology, Chennai, Tamil Nadu India.
2A.Jackulin Mahariba, SRM Institute of science and Technology, Tamil Nadu India.
Manuscript received on 15 September 2019 | Revised Manuscript received on 23 September 2019 | Manuscript Published on 11 October 2019 | PP: 1050-1054 | Volume-8 Issue-11S September 2019 | Retrieval Number: K121609811S19/2019©BEIESP | DOI: 10.35940/ijitee.K1216.09811S19
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Air passengers prediction is said to be the centre of gravity of the growth. With people on the move constantly, there is bound to be some dissatisfaction amongst the customers which could be due to various reason, varying from overbooking of flights to ground operations. This dissatisfaction can be controlled till a limit, in ballpark figuring. In the past, this has been done using various machine learning techniques. For this prediction, in this project, ARIMA Modeling is used which is a time series forecasting method, based on machine learning. To test the stationarity of the data, which is done using Dickey Fuller test. If the data is stationary, it is fit into the ARIMA Model. If the data isn’t stationary, it is made stationary by differencing or by logarithmic transformation. The logarithmic method to make the data stationary. Once the data is stationary, using the Partial autocorrelation function and the autocorrelation function, values of p and q are found, which are required in the time series method. These values are then fit into the ARIMA Modeling and hence, the results are predicted. Upon the use and fitting of various models, the ARIMA(2,1,2) has been the best fit, having the least RMS and RMSE values.
Keywords: ARIMA, Forecasting, Air Passengers, Machine learning
Scope of the Article: Machine learning