The Prediction of Flight Delays Using Regression Method

--EE608 Final Project

INTRODUCTION

Background

Operating an air transportation system efficiently is a highly complex problem.

However, flight delays are very important to both passengers and airlines. Passengers will feel happy and satisfied if their flight take off and land on time. Airlines can save money because reducing the average delay per flight by one minute can save mid-sized airlines millions of dollars in annual crew costs and fuel savings.

Key word: Flight Delay Prediction, Data analysis, Regression

Problem

  • Crowded airspace becoming unpredictable.
  • Rescheduling of critical airspace operations because of delays.
  • Bad customer experience.
  • How to rank different airlines.

OBJECTIVE

  • Explore and analysis the data
  • Delay Prediction
  • Airlines Rank
  • Airlines Recommendation

FUTURE WORK

  1. We would like to further improve our models, perhaps with more training-data or deep neural network.
  2. We will implement prediction results on airline rank.
  3. Airports comparison will be added.
  4. Applied other regularization methods (LASSO…).

TEAM MEMBERS

Ziran Gong

zgong5@stevens.edu

Master of Engineering in AIECE

Yuqing Luo

yluo27@stevens.edu

Master of Science in CPE

Bowen Li

bli50@stevens.edu

Master of Engineering in AIECE

FIND US

OUR LOCATION

Burchard Building

Stevens Institute of Technology

1 Castle Point Terrace

Hoboken, NJ 07030