Research On Support Vector Machine Classification Method Based On Intelligent Optimization

Support vector machine(SVM)has a good performance in solving classification problem,among which feature selection and parameter optimization have a great influence on classification accuracy.There are some redundant features in the feature,which increase the time and space complexity of the algorithm,so the feature selection must be used to reduce the dimension of the data.In addition,the parameters will also affect the final classification accuracy.These two factors influence each other,and how to optimize the feature selection and parameters synchronization to improve the efficiency of classification has become a research trend.Aiming at this problem,an improved algorithm based on particle swarm optimization(PSO)for feature selection and SVM parameters optimization(GPSO-SVM)is proposed to improve the classification accuracy and select the best feature subset.In order to solve the problem that the traditional particle swarm algorithm is easy to fall into local optimum and premature maturation,the algorithm introduces the crossover and mutation operator from genetic algorithm(GA)that allows the particle to carry out cross and mutation operations after iteration and update to avoid the premature problem in PSO.The crossover particle is selected by the irrelevance index between the particles and the mutation probability is determined by the fitness value of the individual,by this way to make the particles jump out of the previous search to the optimal position to improved the diversity of the population and found a better value.Compared with the standard particle swarm algorithm and genetic algorithm for feature selection and parameter optimization,experimental result show that GPSO-SVM can find the appropriate feature subset and parameters of the SVM to obtain better classification performance.Although the GPSO-SVM algorithm improves the accuracy,it has a longer iteration period.This paper uses mind evolutionary algorithm(MEA)to optimize the feature selection and parameter synchronization.Mind evolutionary algorithm is used for feature selection and SVM parameters optimization can achieve good classification results,but there are also some problems such as easy to fall into local optimum.Aiming at this problem,an improved mind evolutionary algorithm for feature selection and SVMparameter optimization is proposed into which the "learning" and "reflection" mechanism are introduced.The algorithm learning through the information sharing among subgroups and reflects through the comparison of the fitness value.In this way to ensure the diversity of population and further improve the classification accuracy.Experimental results show the effectiveness of the algorithm.In this paper,we use two optimization algorithms to solve SVM feature selection and parameter optimization problem,the classification accuracy is higher and the feature selection ability is stronger,and finally achieved good results...