Predictions for Airline Delays

Thanksgiving is one of the most traveled holiday of the year. According to AAA, an estimated 44 million people travel during the thanksgiving holiday and many people choose to travel by plane. That works fine most of the time, until you have a plane delay. We all hate delays and usually no one can see them coming, but recently researchers at Binghamton University, State University of New York (SUNY) have developed a new computer model which they say can predict airline delays faster than anything else in existence. Currently the way these predication models work is by using an artificial neural network. These work by analyzing different data associated with flight delay using a series of different algorithms. All together these algorithms are able to collectively analyze and predict flight delays based on variables like weather or security risks similar to how neurons in our brain work together to solve a problem (the neural in artificial neural network). This new ANN system analyzes 14 of these qualitative variables and tries to give the user a clearer relationship between variables which the system analyzes, and the delays that may or may not have been caused by it. In synthetic tests, this system made predictions that were 20% more accurate and 40% faster than systems currently in use. This could be big for heavy travel days during the year and could even help airports have more flights per day, hopefully reducing the time we have to spend waiting around at an airport. 

Sources:

http://traveltips.usatoday.com/busiest-travel-day-108182.html

https://www.sciencedaily.com/releases/2016/11/161114103905.htm

http://travelsort.com/blog/flight-cancellation-or-delay-tips-for-what-to-do