This project addresses the significant challenge of predicting flight delays using advanced machine learning models, focusing on binary and multi-class classification to ascertain the presence and duration of delays. Utilizing comprehensive datasets, including detailed weather and operational conditions from 2014 to 2018, the analysis identifies key factors such as severe weather, specific airlines, and departure times as critical predictors. These models demonstrate a substantial improvement in forecasting accuracies, offering practical insights for operational adjustments in real-time scenarios that could considerably mitigate disruptions and enhance airline operational efficiency.
Find out more about our data here: Data
View our presentation here: Presentation note: this presentation was given before our project was finished and does not include the final results of our models.
View our final report here: Final Report