Parameters Optimization And Application Research Of V-SVM In Speech Recognition

With the development of the society, all kinds of machines take part in the human production activities and social activities therefore, it is more and more important that to improve the relationship between man and machine that can make the machines work better for human. SVM is a Classifier Machine.it is a new algorithm about Learning Machine. The standard SVM is a universal method of Learning Machine, parameters selection in SVM affect the Learning Machine performance directly, But the existing research on the parameters selection of SVM still have no unified method. So the research of the parameters selection of SVM has great significanceThe selection of the types of kernel function and the parameter values in SVM directly affect the recognition rate in speech recognition system. However, the selection of the types of kernel function and the parameters have not had scientific method, getting the results of the selection usually accord to experience and do a lot of repeated experiments,with great limitations. In order to verify the recognition effect of support vector machine in speech recognition system, this paper constructed four non-specific person and isolated words speech recognition systems which are based on support vector machines of different kernel function respectively and did a lot of simulation experiments. These four kernel function are linear kernel function, radial basis kernel function, three-order polynomial kernel function and sigmoid kernel function. The experimental results show that the recognition results of speech recognition systems which are based on linear kernel support vector machine, radial basis kernel support vector machine and three-order polynomial support vector machine have better recognition effect and faster training time. But the recognition results of speech recognition system based on sigmoid kernel support vector machine are very bad. So the type of kernel functions directly affects the classification performance of support vector machine and accordingly affects the recognition effect of the speech recognition system.In order to study the influences of the selection of the kernel parameter values in condition of the kernel function type is fixed and avoid the difficult problem of selecting parameter values in C-SVM. This paper makes a further study, This text use the changed SVM which is v-SVM, and select parameters of v-SVM based on PSO, and use the optimized parameters in the Speech Recognition System. The results of the experiment show that the method is effective feasible, the optimized parameters make v-SVM have good generalization.