Comparative Analysis of Machine Learning Random Forest, Naïve Bayes and SVM for Flight Delay Classification

Flight delays significantly impact airlines’ operational, customer satisfaction, and economic efficiencies. This paper’s goal is to compare the performance of three machine learning models: Random Forest (RF), Naïve Bayes (NB), and Support Vector Machine (SVM). The real dataset was obtained from the Bureau of Transportation Statistics (BTS), a division of the United States Department of Transportation, containing flight-related information such as scheduled departure and arrival times, flight cancellations, and delay reasons. The model’s performance was evaluated using accuracy, precision, recall, F1-score, and ROC-AUC metrics. The results indicate that Random Forest was the best model with the highest accuracy 92.99%, precision 92.94%, recall 93.81%, F1-score 93.37%, and ROC-AUC 98.39%, indicating its great overall performance. The SVM model ranked second with an accuracy of 85,81%. Naïve Bayes showed limitations with 56,79% recall score, indicating its limitation in detecting actual delay cases. This study provides insightful information that can be used to prepare better flight scheduling and develop better mitigation strategies.