The proliferation of Distributed Denial of Service (DDoS) attacks poses a significant threat to internet users and organizations. This paper presents a comprehensive evaluation of four machine learning algorithms, Catboost, Logistic Regression, Gradient Boost, and Naïve Bayes aimed at effectively detecting and classifying benign and attack transactions. Utilizing four open-source datasets, including CICDDOS19 (Portmapper attack dataset), DDOS Backscatter TCP dataset, DDOS Backscatter ICMP attack dataset, and the SDN dataset sourced from publicly available repositories, meticulous preprocessing techniques were employed to clean and scale the datasets before training the machine learning models. The performance evaluation of the algorithms was conducted based on key metrics such as Accuracy, F1 score, Precision, and Recall. The comparative analysis presented in this paper facilitates the determination of the best-performing algorithm for DDoS attack detection across diverse datasets, thereby offering valuable insights for enhancing cybersecurity measures.