Is this project an undergraduate, graduate, or faculty project?

Undergraduate

Project Type

group

Campus

Daytona Beach

Authors' Class Standing

Taryn Trimble, Senior Ioannis Paraschos, Junior Eshna Bhargava, Junior

Lead Presenter's Name

Taryn Trimble

Lead Presenter's College

DB College of Arts and Sciences

Faculty Mentor Name

Mihhail Berezovski

Abstract

Partnering with OneSky Flight LLC, this project aims to develop a flight time predictor using various machine learning methods. OneSky Flight LLC's goal is to provide IT support for various private jet companies. Their flight time predictor is mainly used for customer satisfaction and aircraft turn-over efficiency. This project involves creating a potential improvement on the current flight time predictor they use now. Six months of flight data was provided by OneSky Flight LLC; it included attributes such as origin, destination, aircraft type, departure time, and arrival time. The two primary methods tested were neural networks and decision trees. Each method was tested with varying architectures and data structures to determine accuracy. The resulting analyses of the architectures found the XGBoost decision tree to be the highest performing model. Using the results of the architectures, an ensemble model can be developed that incorporates the use of both neural networks and decision trees to further increase the accuracy of the predictor.

Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?

Yes, Spark Grant

Share

COinS
 

Predicting Flight Time Using Machine Learning Methods

Partnering with OneSky Flight LLC, this project aims to develop a flight time predictor using various machine learning methods. OneSky Flight LLC's goal is to provide IT support for various private jet companies. Their flight time predictor is mainly used for customer satisfaction and aircraft turn-over efficiency. This project involves creating a potential improvement on the current flight time predictor they use now. Six months of flight data was provided by OneSky Flight LLC; it included attributes such as origin, destination, aircraft type, departure time, and arrival time. The two primary methods tested were neural networks and decision trees. Each method was tested with varying architectures and data structures to determine accuracy. The resulting analyses of the architectures found the XGBoost decision tree to be the highest performing model. Using the results of the architectures, an ensemble model can be developed that incorporates the use of both neural networks and decision trees to further increase the accuracy of the predictor.

 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.