Uçuş gecikmelerini ve iptallerini tahmin etmek için hibrit tahmini modelleme yaklaşımı

This thesis aims to address the critical challenges in the aviation industry by developing predictive models for optimizing flight operations. The scope encompasses predicting flight delays and cancellations. The research investigates various data-driven methods, including machine learning and time series analysis, to improve airline resource management, enhance passenger satisfaction, and reduce operational costs. Method: The study employs a comprehensive methodology involving data collection, preprocessing, feature engineering, and the exploration of diverse predictive modeling techniques. Machine learning algorithms, time series analysis, and regression models will be examined for their applicability to different aspects of flight predictions. Real-world airline data will be used to evaluate the performance and suitability of these methods. Original Value: This research introduces an inclusive approach to predictive modeling for flight delays and cancellations. By focusing on enhancing the accuracy of flight delay predictions, mitigating cancellations, it aims to offer practical solutions for the aviation industry. The originality of this work lies in its adaptability to explore various methods, identify the most effective ones, and present valuable insights for resource allocation and customer service enhancement. Expected Results: The expected results encompass the development of predictive models with high accuracy, addressing the specific challenges of the airline industry. These models should lead to improved resource management, reduced operational costs, and elevated passenger satisfaction. The research is anticipated to offer valuable insights and recommendations for airlines to optimize their flight operations, contributing to a more reliable and customer-centric industry.