Aggressive driving behaviour is a critical issue with significant implications for road safety and traffic management. Researchers have extensively studied methods to predict such behaviours, utilizing various data sources and machine learning techniques. This research focuses on predicting aggressive driving behaviour using GPS data from the Green University Bus Route Dataset. We developed a novel preprocessing algorithm, along with label encoding and standardization techniques, to prepare the dataset for analysis. Seven machine-learning algorithms were used to guess how often aggressive driving would happen. These were Decision Tree classifiers (DTC), AdaBoost classifiers (ABC), Gradient Boosting classifiers (GBC), Extra Tree classifiers (ETC), Support Vector classifiers (SVC), Gaussian Naive Bayes (GNB), and Random Forest classifiers (RFC). Our results indicate that the GBC, ETC, SVC, and RFC outperform other models, with each one achieving the highest accuracy of 99.445\%. This study demonstrates the potential of GPS data to effectively predict aggressive driving behaviour, offering valuable insights for improving road safety measures.