Support Vector Machine Classification Algorithm based on Improved Particle Swarm Optimization

In order to solve the problem that the penalty factor of the main parameters of support vector machine and the selection of kernel parameter has a great influence on the classification accuracy, a support vector machine classification algorithm with improved particle swarm optimization(IPSO-SVM) is proposed. In order to avoid the defects that the standard particle swarm optimization is easy to fall into the local optimal value and the late oscillation convergence speed is slow, the algorithm introduces the Gaussian nonlinear decreasing inertia weight and asynchronous learning factor change strategy into the standard particle swarm optimization to optimize the support vector machine parameters, so as to improve the optimization ability. Five groups of common data sets are used for experiments. Compared with support vector machine classification algorithm and support vector machine classification algorithm based on genetic algorithm (GA-SVM), the best classification accuracy of IPSO-SVM is the highest, the average classification accuracy is improved by more than 1.5%, and the classification time is short, which verifies that the algorithm has a good classification effect.