A Novel Approach for Statistical Analysis of CPU Utilization in Cloud Computing using Graphic Clustering Algorithm with Hierarchical Clustering Algorithm based on Accuracy and Root Mean Square Error (RSME) Value

Abstract

Cloud computing promotes resource sharing by lowering the hardware cost for business users, while also promising improved energy efficiency and resource management. To the service providers a significant statistic is cpu utilization. To avoid the performance issues to keep track of cpu utilization and manage resource usage. Materials and Methods: Classification of suspicious dataset using Graph Clustering Algorithm (N=4) and Hierarchical Clustering Algorithm (N =4) with feature extraction which help s for better performance of accuracy with statistical significance between Graph Clustering Algorithm and Hierarchical Clustering Algorithm (p=0.080%). Results: The accuracy of abnormal event detection using Graph Clustering Algorithm model is 98.7% and Hierarchical Clustering Algorithm is 90.1 % significant difference between Graph Clustering Algorithm and Hierarchical Clustering Algorithm is (p<0.01) Graph clustering Algorithm performs (98%) better than Hierarchical Clustering Algorithm (90%). Conclusion: The proposed Graph Clustering Algorithm and Hierarchical Clustering Algorithm seems to be significantly better in predicting datasets with more accuracy than the hierarchical clustering algorithm.