Research On Server Performance Analysis And Prediction Based On SVM Algorithms

With the continuous development of the Internet,Web services have penetrated into people's lives,and the performance of Web servers directly determines the advantages and disadvantages of Web services.Therefore,it is particularly important to accurately predict the performance of Web servers.In order to improve the accuracy of Web server performance prediction,this paper establishes an accurate server performance prediction model based on Intelligent algorithm,and puts forward an improved idea for the classical algorithm.The main work and research results are as follows:1.The basic concepts and research status of server performance are briefly described and explained.The algorithms and models used to predict server performance at home and abroad are compared,and their advantages and disadvantages are analyzed.Aiming at the prediction of server performance,this paper focuses on the support vector machine(SVM)algorithm,which lays a theoretical foundation for the follow-up work of this paper.2.Server performance prediction focuses on the SVM model algorithm.To solve the problems of slow prediction speed and large deviation of prediction results,the gray wolf optimization algorithm(GWO)is used to optimize the model.However,in the face of complex optimization problems,GWO algorithm has low precision,unstable convergence state and random initial population generation.In this paper,DE algorithm is used to generate more suitable initial population for gray wolf optimization algorithm.The improved model is DE-GWO-SVM model.3.In this paper,a second prediction method,LS-SVM algorithm based on time series,is proposed for server performance prediction.However,the prediction accuracy of the model is low and the prediction efficiency is slow.It is necessary to select appropriate kernel function parameters ? and penalty factor C.In this paper,the improved PSO algorithm is used to find the optimal parameters ? and C of LS-SVM model.The improved PSO' LS-SVM algorithm is used to predict the performance of time series-based servers.The improved model is IMP-PSO-LSSVMmodel.In this paper,the commonly used quality evaluation methods are used to evaluate the prediction model.The results show that the two improved models have accurate prediction effect,which further proves the rationality and effectiveness of the improved algorithm.