A Review on Machine Learning Methods for Workload Prediction in Cloud Computing

Workload prediction is one of the critical parts of resource provisioning in cloud computing and its evolved branches such as serverless and edge computing. Effective resource provisioning stands as a crucial element within the realm of edge-cloud computing. Accurate prediction of cloud workloads is essential for the effective allocation of resources. Workload prediction plays a crucial role in enhancing efficiency, reducing costs, optimizing cloud performance, maintaining a high level of quality of service, and minimizing energy consumption. In this paper, we conduct a comprehensive review of state-of-the-art Machine Learning (ML) and Deep Learning (DL) algorithms employed in workload prediction in cloud computing and other similar platforms such as edge computing. We compared the selected papers in terms of utilized methods and techniques, predicted factors, accuracy metrics, and the dataset. Additionally, to facilitate usability and comparison, articles sharing similar advantages and disadvantages are organized into a table. Finally, the paper concludes by addressing current challenges and future research directions.