Flight Delays in the US

Data on US Flights was used to train an ensamble model to predict flight delays from multiple carriers.

Predicting Airport Delays is a project in which we trained multiple classification models to predict if a flight would be delayed given the weather and carriers information two hours before the flight. This work was the final project of the course Machine Learning at Scale. Authors: Ruth Ashford, Spencer Song, Rajiv Verma, Lana Elauria, Carolina Arriaga.

Objective

We will be predicting which flights will be delayed at departure by more than 15 minutes. We will be making the prediction two hours in advance to the scheduled departure time. This will benefit airlines who are behind in On-Time Performance or OTP (OTP is defined as the percentage of flights having less than 15 mins of delay) for proactive cost-savings.


© 2019. All rights reserved.