Forecasting air passenger volume in Singapore : an evaluation of time series models and econometric models

Nowadays due to the increasingly development of air transport technology, air passenger movements has been growing dynamically. In addition, Singapore aviation industry contributes a large part to Singapore economy. However, there is a high possibility that Singapore may face dilemma for air traffic...

Full description

Saved in:
Bibliographic Details
Main Author: Guo, Rui
Other Authors: Zhong Zhaowei
Format: Theses and Dissertations
Language:English
Published: 2017
Subjects:
Online Access:http://hdl.handle.net/10356/69907
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Nowadays due to the increasingly development of air transport technology, air passenger movements has been growing dynamically. In addition, Singapore aviation industry contributes a large part to Singapore economy. However, there is a high possibility that Singapore may face dilemma for air traffic congestion and service standard reduction without proper estimation of air passenger volume growth. To prevent such issue, it is necessary to have a good forecasting model suitable for Singapore situation. This project explores various methods to predict the air passenger movements and analyzes and compares the relative results from corresponding models. 13 models inclusive of time-series models and econometric models were simulated for 18 years prediction from 1998 to 2015 in the report and compared with each other using error measurement. MAPE (mean absolute percentage error), RMSE (root mean square error), absolute value of largest degree of divergence, and Dbal are utilized for performance gauge. Finally, appropriate models for Singapore situation is to be recommended, which are quadratic trend and ARIMA model in this report. Simultaneously, different variables, such as Singapore GDP, China GDP and exchange rate and so on, were tried in econometric models. The most appropriate variables were chosen. Afterwards, forecasting for the next 18 years is conducted by using quadratic trend model, ARIMA model and one econometric model.