Cloud computing is the need of the day for computing but vulnerable to cyber-attacks, particularly those presented by Distributed Denial of Service or DDoS attacks, which essentially crash services through heavy-handed levels of attack traffic. Traditional methods of detection are not dynamic enough for cloud environments, necessitating the evolution towards complex solutions such as machine learning (ML) for attack identification. This paper elaborates on three algorithms: Decision Tree, Naïve Bayes, and Support Vector Machine SVM in identifying DDoS attacks using the CICDDoS2019 dataset. The results of the study were based on accuracy, precision, recall, and the computational cost for each model considered. It seems that the Decision Tree model is accurate up to 100% while accuracy in Naïve Bayes and SVM stands at 99.93%. However, every algorithm has an accompanying limitation such as the overfitting of Decision Trees and high computational requirements of SVM. In this paper, the adoption of hybrid machine learning models and feature engineering can potentially improve DDoS detection for reliability and robustness of cloud system.