Malik Adeel Khokhar, Noor-ul-Ain and Laiba Arshad2025-10-022025-10-022024https://escholar.umt.edu.pk/handle/123456789/7789Flight delays pose a significant challenge for the aviation industry, impacting airlines, airports, air traffic controllers, facility managers, and passengers. The development of a highly accurate prediction model is the need of the hour to enable informed decision-making. This paper proposes a model based on flight data attributes and adverse weather conditions, utilizing data mining and supervised machine learning algorithms. Deep machine learning algorithms and neural network models were applied to make accurate predictions. Each model's prediction accuracy and receiver operating characteristic (ROC) curves were compared to evaluate their performance. This model was developed in Pakistan at three major airports, including Islamabad (ISB), Lahore (LHE), and Karachi (KHI). The ultimate objective of the project is to address the significant shortcomings in the region's capability to predict aircraft delays with precision. This study is competitive as previous models cover one airport, but it covers three. For the prediction phase, flight attributes and weather data were input into the model. The trained model will then predict whether a flight will be on time or delayed based on specific groupings. The model achieved a mean absolute error of 23, which means the accuracy rate was remarkable, given the specific key attributes of data and weather. This proposed model aims to boost operational efficiency and passenger satisfaction by further enhancement.enThesis