Predicting Flight Delays using Advanced ML Algorithms

Date of Award


Document Type


Degree Name

Master of Science in Machine Learning


Machine Learning

First Advisor

Dr. Zhiqiang Xu

Second Advisor

Dr. Bin Gu


This thesis provides an in-depth exploration and analysis of the meteoric rise in Emirate’s civil aviation sector, focusing on the period from 2006 to 2017. It delves into how the burgeoning growth in passenger traffic and aviation infrastructure has influenced the region’s economic landscape, with a particular emphasis on Dubai International Airport’s expanding role as a global aviation hub. The research critically examines the escalation in passenger numbers and turnover volumes, linking this growth to broader economic indicators and strategic infrastructural developments. Moreover, the thesis addresses the persistent challenge of flight delays within the aviation industry, identifying and analyzing the multifaceted factors contributing to these disruptions. By investigating the interplay of adverse weather conditions, airline scheduling, air traffic control efficiencies, and other pivotal factors, the study sheds light on the complexities surrounding flight delay causation and its consequential impacts on airlines, airports, and passengers. Through a methodical literature review and a detailed analysis of flight delay prediction methodologies, the research identifies existing gaps in academic literature and practical applications, proposing a comprehensive analytical framework to enhance the predictive accuracy and operational utility of delay forecasting models. The study’s findings aim to contribute significantly to the body of knowledge in aviation research, offering actionable insights for improving air traffic management and enhancing the overall efficiency and reliability of the aviation sector in the Emirates. In addressing the challenge of flight delay prediction, this thesis introduces a customdeveloped Artificial Neural Network (ANN) model, meticulously engineered to dissect and forecast the intricate patterns of flight delays. The model’s architecture is thoughtfully designed with multiple layers and neurons, tailored to capture the nuanced dependencies and variables influencing delay occurrences in the aviation sector. Utilizing a deep learning framework, the ANN model demonstrates exceptional proficiency in analyzing historical f light data, identifying key delay determinants, and predicting future incidents with remarkable accuracy. The model’s predictive capabilities are rigorously validated against established metrics, showcasing its potential to significantly enhance operational decisionmaking and strategic planning in civil aviation. Through this innovative approach, the research elucidates the transformative impact of deep learning in advancing the predictive analytics domain within the aviation industry, offering a robust tool for mitigating the pervasive challenge of flight delays.


Thesis submitted to the Deanship of Graduate and Postdoctoral Studies

In partial fulfilment of the requirements for the M.Sc degree in Machine Learning

Advisors:Zhiqiang Xu, Bin Gu

Online access available for MBZUAI patrons