A Deep BiLSTM Machine Learning Method for Flight Delay Prediction Classification

This paper proposes a classification approach for flight delays using Bidirectional Long Short-Term Memory (BiLSTM) and (LSTM) models. Flight are major issue in the airline industry, causing inconvenience to passengers financial losses airlines. The BiLSTM LSTM models, powerful deep learning techniques, have shown promising results task. In this study, we collected dataset from United States (US) Bureau of Transportation Statistics (BTS) on-time performance information used it train test We set three criteria selecting highly important features evaluation models Confusion matrix shows that outperforms model. evaluating Mathews Correlation Coefficient (MCC), model offers better correlation 0.99 between original predicted classes. Our experiment predicting delays, takes advantage forward backward hidden sequences neural network exploration exploitation achieve high accuracy, recall, F1-Score. findings suggest can effectively predict provide valuable airlines, passengers, airport managers.