Beyond the Model: A Deployable Machine Learning System for Flight Delay Prediction Using Comparative Regression Analysis

This study presented a machine learning-based approach for predicting flight arrival delays using historical Indian domestic flight data. The primary objective was to develop an accurate and practical prediction system while conducting a comparative performance analysis of multiple regression algorithms. We used four machine learning models. Ridge Regression, Decision Tree Regression, Random Forest Regression and Extra Trees Regression. We implemented these machine learning models. Then checked how well they worked. We did this by using the steps to prepare the data and the same standards to evaluate the results of Ridge Regression and the other machine learning models. Our approach was very structured. We started by getting the dataset ready. Then we cleaned up the data. This included fixing missing values getting rid of records and making sure the numbers were on the same scale. We also changed the categories into numbers that the machine learning models could understand. After that we split the data into two parts. One for training and one for testing. This way we could make sure that our results, for Ridge Regression and the other machine learning models were fair and accurate. Then we trained the machine learning models. Checked how well they worked. Finally we put the machine learning models into a web application. The comparative evaluation framework demonstrated the importance of assessing multiple algorithms rather than relying on a single predictive technique. The selected regression models provide different learning characteristics, enabling the study to analyze their effectiveness in estimating flight arrival delays under identical experimental conditions. This comparative approach supports the identification of the most appropriate prediction model based on accuracy, reliability, and generalization capability. The developed Flight Delay Prediction System extends beyond model development by integrating the selected prediction model into a FastAPI-based web application. The system enables users to enter flight information, obtain predicted arrival delays, review previous prediction records, and generate downloadable PDF reports. This integration illustrates how machine learning models can be transformed into practical decision-support tools that improve user accessibility and operational usability. Overall, the proposed framework demonstrates the applicability of machine learning for flight delay prediction and provides a systematic methodology for comparing predictive models using historical aviation data. The study contributes to the growing use of predictive analytics in the aviation sector by combining comparative model evaluation with a user-oriented software implementation, thereby supporting improved travel planning and informed operational decision-making.

[1] B. Mtimkulu and B. Maphosa, "Flight Delay Prediction Using Machine Learning: A Comparative Study of Ensemble Techniques," Proc. International Conference on Computing and Data Analytics, 2023. [2] A. Jha, et al., "Flight Delay Prediction Using Hybrid Machine Learning Approach: A Case Study of Major Airlines in the United States," International Journal of Intelligent Systems and Applications, vol. 16, no. 2, pp. 45–61, 2024. [3] M. Hatıpoğlu and E. Tosun, "Predictive Modeling of Flight Delays at an Airport Using Machine Learning Methods," Applied Sciences, vol. 14, no. 3, 2024. [4] A. AlBassam and M. AlShahrani, "Flight Delay Prediction: Evaluating Machine Learning Algorithms for Enhanced Accuracy," PLOS ONE, vol. 20, no. 1, 2025. [5] Y. Dai, "A Hybrid Machine Learning Model for Flight Delay Prediction Using Aviation Big Data," Scientific Reports, vol. 14, 2024. [6] G. Gui, et al., "Flight Delay Prediction Based on Aviation Big Data and Machine Learning," IEEE Access, vol. 8, pp. 197432–197444, 2020. [7] Y. Zhong, et al., "FlightNet-ST: An Interpretable Spatio-Temporal Deep Learning Framework for Civil Aviation Flight Delay Prediction," Expert Systems with Applications, 2025. [8] R. Kothari, R. Kakkar, S. Agrawal, P. Oza, S. Tanwar, et al., "Selection of Best Machine Learning Model to Predict Delay in Passenger Airlines," IEEE Access, vol. 11, pp. 79673–79689, 2023. [9] J. Rebollo and H. Balakrishnan, "Characterization and Prediction of Air Traffic Delays," Transportation Research Part C: Emerging Technologies, vol. 44, pp. 231–241, 2014. [10] J. Qu, X. Dong, W. Zhang, and Y. Qu, "A Machine Learning Approach for Flight Delay Prediction," in Proc. IEEE International Conference on Big Data Analysis (ICBDA), 2016.