AN EXPERIMENTAL STUDY OF FLIGHT DELAY PREDICTION WITH BIG DATA

MPhil Thesis Defence


Title: "AN EXPERIMENTAL STUDY OF FLIGHT DELAY PREDICTION WITH BIG DATA"

By

Miss Jingshu PENG


Abstract

Flight delays happen frequently due to various reasons such as adverse weather 
at airports, shortage of airport runway capacity, the increase in the number of 
aircrafts and poor air traffic control. For example, busy mainland China 
airports, such as Pudong Airport in Shanghai, were reported as the worst of the 
world with punctual departure rate at only 37.26 percent. This results in 
challenge of schedule planning for passengers such that different links of the 
flight schedule connect well and users are satisfied with the traveling plan 
(e.g., enough connection time, small traveling time of the trip, etc.). To 
address this problem, we propose a flight delay prediction system which 
consists of three components: 1) data cleaning/integration, 2) feature 
engineering/classifiers, and 3) delay prediction model. We show that complete 
aviation big data can make flight delay prediction much easier with higher 
accuracy and better performance, confirmed by experimental results which reveal 
that the prediction performance has been significantly improve by using a large 
volume of training data and a large variety of features. Potentially, the 
unsophisticatedness of migrating to other aviation databases leads to enormous 
business value.


Date:			Friday, 12 August 2016

Time:			9:00am - 11:00am

Venue:			Room 3494
 			Lifts 25/26

Committee Members:	Prof. Lionel Ni (Supervisor)
 			Dr. Qiong Luo (Chairperson)
 			Dr. Yangqiu Song


**** ALL are Welcome ****