A Comparative Review of AI-Driven Predictive Resource Management for Cost, Performance, and Energy Optimization in Cloud Computing

This paper reviewed and compared AI-driven approaches for predictive workload management, dynamic resource allocation, autoscaling, cost optimization, multi-cloud coordination, and energy-efficient cloud operation. The literature consistently indicates that AI and ML can move resource management from reactive provisioning toward proactive decision-making. Workload prediction provides the information needed for anticipation, while reinforcement learning, optimization, and policy-based mechanisms can convert predictions into allocation and scaling actions. The main research gap is the absence of a consistently evaluated, integrated framework that simultaneously considers prediction accuracy, cost, performance, SLA compliance, resource utilization, and energy or carbon impact. Future work should therefore focus on benchmark-driven experiments using common datasets, multiple ML baselines, standardized metrics, multi-objective optimization, online adaptation, and real-world cloud validation. Such a framework can provide stronger evidence for practical deployment of intelligent and sustainable cloud resource management.

[1] A. Adeyinka, “Dynamic resource allocation using AI-driven workload forecasting in multi-cloud environments,” World Journal of Advanced Research and Reviews, vol. 23, no. 1, pp. 3188–3198, 2024, doi: 10.30574/wjarr.2024.23.1.2178. [2] V. S. Machiraju, V. Kumar, and S. Sharma, “ML-Based Autoscaling for Elastic Cloud Applications: Taxonomy, Frameworks, and Evaluation,” Mathematical and Computational Applications, vol. 31, no. 2, art. 49, 2026, doi: 10.3390/mca31020049. [3] J. Kumar and A. K. Singh, “Workload prediction in cloud using artificial neural network and adaptive differential evolution,” Future Generation Computer Systems, vol. 81, pp. 41–52, 2018, doi: 10.1016/j.future.2017.10.047. [4] D. Bodra and S. Khairnar, “Machine learning-based cloud resource allocation algorithms: a comprehensive comparative review,” Frontiers in Computer Science, vol. 7, art. 1678976, 2025, doi: 10.3389/fcomp.2025.1678976. [5] F. Haroon, “Intelligent Resource Allocation in Cloud Computing Environments: Leveraging Machine Learning for Dynamic Workload Balancing, Cost Efficiency, and Performance Optimization,” Annual Methodological Archive Research Review, vol. 3, no. 4, pp. 526–549, 2025, doi: 10.63075/gdxsyd86. [6] R. Yadav and J. S. Yadav, “A Comprehensive Review of Machine Learning Approaches for Predictive Resource Management in Cloud Computing Environment,” International Journal of Scientific Research in Computer Science, Engineering and Information Technology, vol. 11, no. 6, pp. 347–354, 2025, doi: 10.32628/CSEIT2511656. [7] Thafzy V. M., “A Review on the Integration of Machine Learning in Cloud Computing Resource Management,” International Journal of Scientific Research and Technology, vol. 2, no. 1, pp. 18–21, 2025, doi: 10.5281/zenodo.14592545. [8] V. Ramamoorthi, “AI-Driven Resource Management in Cloud Computing: A Review,” International Journal of Scientific Research in Science and Technology, vol. 11, no. 2, pp. 683–696, 2024, doi: 10.32628/IJSRST24113260. [9] S. K. Choudhary, “AI-Powered Predictive Analytics for Dynamic Cloud Resource Optimization: A Technical Implementation Framework,” International Journal of Scientific Research in Computer Science, Engineering and Information Technology, vol. 11, no. 1, pp. 1267–1275, 2025, doi: 10.32628/CSEIT251112122.  [10] V. Punniyamoorthy, A. K. Agarwal, B. Kumar, A. Mazumder, K. Kannan, and S. Saha, “AI-Driven Cloud Resource Optimization for Multi-Cluster Environments,” arXiv, 2025, arXiv:2512.24914. [11] O. R. Polu, “AI-Enhanced Cloud Cost Optimization Using Predictive Analytics,” International Journal of Artificial Intelligence Research and Development, vol. 3, no. 1, pp. 74–88, 2025, doi: 10.34218/IJAIRD_03_01_006.