Exploring Machine Learning Algorithms to Enhance Cloud Comput‑ ing Security

The increasing adoption of cloud computing (CC) has introduced significant security and privacy concerns, demanding intelligent and adaptive solutions. This review explores the application of machine learning (ML) algorithms—both supervised and unsupervised—in addressing these challenges within cloud environments. A total of 87 peer‑reviewed studies published between 2014 and 2025 were analyzed to assess the effectiveness of various ML techniques. Supervised Machine Learning (SML) algorithms such as Artificial Neural Networks (ANNs), Support Vector Machines (SVM), K‑Nearest Neighbors (K‑NN), Naive Bayes, and C4.5 Decision Trees are examined for their effectiveness in intrusion detection, anomaly classification, and threat mitigation. Concurrently, Unsupervised Machine Learning (UML) algorithms, including Unsupervised Neural Networks (UNNs), K‑Means clustering, and Singular Value Decomposition (SVD), are analyzed for their capacity to detect unknown threats and extract latent patterns from unlabeled data. Key trends reveal a growing preference for hybrid models, the superior accuracy of deep learning in anomaly detection, and the emerging use of context‑aware frameworks. The review shows a comparative analysis of these approaches, highlighting their advantages, limitations, and application scenarios in cloud security. Future research directions are proposed, emphasizing hybrid learning models, enhanced datasets, and context‑aware security frameworks. The findings underscore the transformative potential of ML in fortifying cloud infrastructures against evolving cyber threats.