Enhancing Flight Delay Prediction and Classification Using a Hybrid Bi-LSTM: Machine Learning

Abstract

The flight delay prediction system plays a vital role in the airlines which helps to predict the delay of flight in real time. Airlines must estimate flight delays accurately because the findings can be used to boost customer satisfaction and airline agency profits. This research introduces a flight delay prediction framework employing the Bidirectional Long Short-Time Memory (Bi-LSTM) model. To capture the long-term dependencies in the sequential data LSTM works as a hybrid component. Overcoming the difficulties faced by the airline industries mainly focusing on flight on-time performance and the weather data through the confusion Matrix The results show that the BiLSTM model improves the LSTM model and achieves the maximum accuracy. The BiLSTM model effectively utilizes both forward and backward hidden sequences. The outcomes demonstrate the Hybrid BiLSTM approach's potential as a useful instrument for enhancing flight delay management in the aviation sector.