Font Size: a A A

Parameters Optimization Of Support Vector Machine Based On Improved Quantum Particle Swarm Optimization Algorithm

Posted on:2016-07-18Degree:MasterType:Thesis
Country:ChinaCandidate:F LiFull Text:PDF
GTID:2348330536987042Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the rapid development of the digitization of information,human beings are entering the era of big data.Big data analysis has been applied in the electronic commerce,financial securities,medical and health industries,which has brought us huge economic and social benefits.Classification,an important data mining analysis method,is conducive to a better overall understanding of the data.Support vector machine(SVM)is widely applied to the classification problem but the parameters of SVM have great influences on classification accuracy.So it is necessary to optimize the parameters of SVM for different classification problems.To enhance the classification accuracy of SVM,this thesis studies the different optimization algorithms to explore the best optimization algorithm for SVM,and eventually promotes the research on theory and application of big data analysis.There are obvious advantages in swarm intelligence algorithm,the main algorithm of SVM,based on the analysis of the advantages and disadvantages of the common optimization algorithms such as the grid search algorithm,the gradient descent method and the swarm intelligence algorithm.Secondly,the classification results of SVM optimized by the genetic algorithm(GA),particle swarm optimization algorithm(PSO)and quantum particle swarm optimization algorithm(QPSO)respectively are analyzed and compared in details,It is shown that SVM optimized by PSO and QPSO is better.Finally,an improved QPSO named double center quantum particle swarm optimization algorithm(DCQPSO)is proposed based on the double center idea.The simulation experiment shows that the improved QPSO is more advantageous.
Keywords/Search Tags:Support Vector Machine, Classification Problem, Grid Search Algorithm, Particle Swarm Optimization Algorithm, Quantum Particle Swarm Optimization Algorithm
PDF Full Text Request
Related items