ShortVIX: Predicting on-time performance of airlines in the United States

Find us online:


Git repo:

Our Inspiration

In the past decade, the United States has seen a large growth of low-cost (LCC) and ultra-low-cost carriers (ULCC) into the commercial aviation business. The Journal of Air Transport Management found that in 2015 when a LCC or ULCC enters a market, airfares in that market decrease by 8% and 21% respectively. And with the introduction of fuel-efficient planes such as the Boeing 737 MAX family and,4 soon, the Airbus 220, airlines will be able to serve passengers who normally do not fly and also operate in markets that previously had little demand. As flights and passengers increase in the US, airports get more crowded leading to higher potential delays. ShortVIX hopes to be the solution by providing passengers with on-time reports of their flight in order for them to ensure that they will have a hassle-free journey.

Our Vision

ShortVIX will create a machine learning model to predict flight on-time performance. Our goal is to gather support data of a flight in order to be determine on-time performance. ShortVIX will collect data on airlines, aircraft type, aircraft age, flight departure and arrival times, departure and arrival airport, buffer time between flight and previous flight, hubs ( in quantiles), airport size, number of flights at airports, and whether the flight takes place during the holiday period.

Our End Goal

ShortVIX hopes to be able to create a predictive aviation analytics model that would be able to help passengers, airlines, and airports predict the probability of their flight being on-time. Therefore, passenger, airlines, and airports can better account for these when they plan schedules their flights.