Authors: Muhammad Tahir
Abstract: Efficient resource management is a critical challenge in cloud computing due to the dynamic and heterogeneous nature of workloads. Machine learning (ML) has emerged as a powerful approach for optimizing cloud resource allocation, utilization, and performance. This study explores the application of ML techniques to enhance cloud resource optimization by enabling predictive, adaptive, and autonomous management strategies. It examines how supervised, unsupervised, and reinforcement learning models can be used to forecast workload demand, perform dynamic resource provisioning, and improve scheduling decisions. The paper also highlights the role of ML in optimizing energy consumption, reducing operational costs, and maintaining quality of service (QoS) in large-scale cloud environments. Key techniques such as workload prediction, anomaly detection, auto-scaling, and intelligent load balancing are discussed within the context of cloud infrastructure. Additionally, the study addresses challenges including data variability, model accuracy, latency, and integration complexity, along with emerging solutions such as edge computing and real-time analytics. The findings emphasize that the integration of machine learning into cloud resource management systems significantly enhances efficiency, scalability, and reliability, making it a vital component of next-generation cloud platforms.
DOI: https://doi.org/10.5281/zenodo.20280054
