Improved Methods for Predicting Flight Delay using Machine Learning Techniques

In recent years, flight delays have been an ongoing obstacle in the aviation sector due to increased air traffic congestion worldwide. Moreover, it also prolongs the other flights, which can discourage users from traveling with the particular airline. Hence, the project uses a machine learning system to predict flight delays based on a data set provided by Kaggle. The attributes of this data set include departure and arrival times, weather conditions, carrier details, airline and airport data, all of which need to be preprocessed through data cleaning, feature engineering, and scaling. There are also several developed predictive models, such as Random Forest, k-Nearest Neighbors, Logistic Regression, and Naive Bayes. The framework includes an interactive web interface constructed with HTML, CSS, and Flask that enables users to submit flight data and retrieve real-time predictions. An Exploratory Data Analysis (EDA) is carried out on the significant factors delaying flights, and the models are hyperparameter-tuned to ensure their optimum deployment.