- Wednesday, Oct 29th, 2025
International Journal of Advanced Research in Arts, Science, Engineering & Management (IJARASEM)
(A High Impact Factor, Monthly, Peer Reviewed Journal)
(A High Impact Factor, Monthly, Peer Reviewed Journal)
| TITLE | A Predictive Analytics Approach to Flight Cancellations using Random Forest |
|---|---|
| ABSTRACT | Flight cancellations pose significant challenges to both passengers and airlines, resulting in financial losses, operational disruptions, and customer dissatisfaction. Accurate and timely prediction of flight cancellations can enable airlines to take proactive measures, optimize resource allocation, and improve passenger experience. This paper presents a machine learning framework for predicting flight cancellations using a Random Forest classifier combined with Synthetic Minority Over-sampling Technique (SMOTE) to address the inherent class imbalance in flight cancellation datasets. The model leverages a rich set of features including temporal attributes (flight date, departure time), geographic information (origin and destination coordinates), and operational details (airline, flight distance, scheduled elapsed time). Extensive feature engineering is performed to extract meaningful variables such as weekend indicators and seasonal categories. The dataset is pre-processed to handle missing values and categorical variables are encoded using label encoding. SMOTE is applied to the training data to synthetically balance the minority class of cancelled flights, improving the model’s sensitivity. The Random Forest classifier is trained and evaluated on a hold-out test set, achieving high recall and F1-score for the cancellation class, demonstrating its effectiveness in identifying flights likely to be cancelled. The paper also details the implementation of an interactive prediction interface that validates user inputs and outputs cancellation probabilities, facilitating practical deployment. Comparative analysis with baseline models highlights the superiority of the proposed approach. Finally, the study discusses limitations and outlines future directions including integration of real-time weather data and exploration of deep learning techniques. This work contributes a robust, interpretable, and scalable solution for flight cancellation prediction, with potential benefits for airline operations and passenger management. |
| AUTHOR | A. Nandhini, Aiswarya.M.P Assistant professor-SG, Department of Computer Applications, Nehru College of Management, Coimbatore, Tamil Nadu, India Student of II MCA, Department of Computer Applications, Nehru College of Management, Coimbatore, Tamil Nadu, India |
| PUBLICATION DATE | 2025-10-27 |
| VOLUME | 12 |
| ISSUE | 5 |
| 17_A Predictive Analytics Approach to Flight Cancellations using Random Forest.pdf | |
| KEYWORDS |
Copyright@IJARASEM