Flight Delay Prediction

      Our task is to predict the arrival delay for US domestic flights based on attributes of the flight, along with its originating and destination airports. Flight delay is annoying as it often disrupts busy schedules and leads to stress and anxiety. When people book flights, they might not only consider the price and timing of the flight, but also hope to avoid possible delays. However, it is difficult to make this judgment by just looking at the airlines, airports, time of the flight etc. Therefore, we want to use ML techniques to build a model to predict whether a specific flight will delay by training past on-time performance data to help passengers make more informed decisions when they book their travels.

      According to Bureau of Transportation Statistics, over 20% of domestic flights was delayed during the past year. Among the delayed flight, around 70% was caused by non-weather reasons, such as air carrier delay, aircraft arriving late and national aviation system delay. In this project, we will specifically focus on using flight information as input features to predict delays that are not caused by extreme weather conditions.