.

ISSN 2063-5346
For urgent queries please contact : +918130348310

IMPROVED ACCURACY IN PREDICTING THE NETWORK PERFORMANCE FOR ALLOCATING SERVER RESOURCES USING LINEAR REGRESSION COMPARED WITH SUPPORT VECTOR REGRESSION.

Main Article Content

Gade. Mary Spandana1, K. Sashi Rekha
» doi: 10.31838/ecb/2023.12.sa1.421

Abstract

Aim: The main objective of this research is to estimate the accuracy rating of predicting network operation for allocating server resources using servers data using the Novel Linear Regression algorithm versus the Support Vector Regression (SVR). Materials and Methods: For trying to predict the accuracy rate of network performance for allocating server resources in information technology infrastructure, at varying times, the Novel Linear Regression algorithm (LR) with sample size=10 and Support Vector Regression (SVR) with sample size=10, 95 percent confidence increment, and pretest power is 80 % were restated. The Novel Linear Regression converts the original into a higher-dimensional space, which enhances accuracy. Results: Novel linear regression appears to prove with better accuracy (88%) compared to Support Vector Regression(78%). There was an insignificant difference between LR and SVR with p=0.950 (p<0.05). Conclusion: When it came to identifying server resources, the Novel Linear Regression algorithm significantly outperforms Support Vector Regression in terms of network performance

Article Details