Leveraging Weather Data for Improved Flight Delay Prediction: A Comparative Analysis of Decision Trees and Random Forests

Suryawanshi, Yash Rajesh (2024) Leveraging Weather Data for Improved Flight Delay Prediction: A Comparative Analysis of Decision Trees and Random Forests. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

Previous research, particularly works that have used models as Random Forests to predict flight delays has shown significant improvement in the prediction accuracy by adding factors like Flight Attributes and Operational Variables. The most frequently considered variables are weather condition that can be determinant in delaying and many of the other models ignore or do not consider them. The studies often use static or simplistic methods to incorporate weather data, not depicting the complex and dynamic system of how operations are affected by the weather acting. For example, a model may use high level categorical variables for weather (e.g. clear, cloudy, rainy) without more detailed info such as wind speed or temperature changes or sudden transitions of the climate which is going to reduce its predictive value.

Furthermore, while a few studies take weather data into account their use of it remains superficial and they do not conduct extensive tests to determine the effectiveness of different machine learning approaches in exploiting this information. The emphasis is commonly on simple predictive performance without considering the reasoning behind feature importance, interpretability and there lacks understanding of which models are better suited to accomplish it.

This study is going to compare the performance of Decision Trees and Random Forests, in searching for model which can help integrate weather variables more effectively with that not only will enhance predictive accuracy but also offer a stronger solution to airlines. Improving efficiency, increasing overall passenger satisfaction and ultimately a comprehensive solution to fight flight delays.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Qayum, Abdul UNSPECIFIED
Uncontrolled Keywords:	Flight Delay Prediction; Machine Learning; Random Forest; Decision Tree; Weather Data Integration; Predictive Modeling; Aviation Operations; Airline Industry; Forecasting Accuracy
Subjects:	G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Aviation Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	26 Aug 2025 11:36
Last Modified:	26 Aug 2025 11:36
URI:	https://norma.ncirl.ie/id/eprint/8641

Actions (login required)

View Item