Review on Machine Learning for Resource Usage Cost Optimization in Cloud Computing

Abstract

Abstract: Small and medium-sized enterprises are increasingly adopting cloud computing, and optimizing the cost of cloud resources has become a crucial concern for them. Although several methods have been proposed to optimize cloud computing resources, these methods mainly focus on a single factor, such as compute power, which may not yield satisfactory results in real-world cloud workloads that are multi-factor, dynamic, and irregular. This paper proposes a new approach that utilizes anomaly detection, machine learning, and particle swarm optimization to achieve a cost-optimal cloud resource configuration.