Distributed Denial-of-Service (DDoS) attacks pose a significant threat to network security, often causing disruptions and substantial losses. Traditional detection techniques struggle to keep pace with evolving attack strategies, necessitating more robust mechanisms. In this thesis, we investigate the application of ensemble learning approaches to enhance DDoS attack detection. Leveraging network traffic features such as UDP, SYN, and DNS, we explore how ensemble learning can improve detection effectiveness and accuracy. By harnessing the strengths of multiple models, ensemble learning offers a promising strategy for developing reliable and widely applicable detection systems. Our study evaluates the effectiveness of ensemble learning techniques, including bagging, boosting, and stacking, in identifying various types of DDoS attacks. Utilizing established DDoS attack datasets, we assess key metrics such as accuracy, precision, recall, and F1-score. Our experiments demonstrate the potential of ensemble learning to significantly enhance the efficiency and accuracy of DDoS attack detection, thereby advancing detection systems.