Research On Classification Algorithm Of Support Vector Machine And Its Applications

Statistical learning theory (SLT) is based on the structural risk minimization (SRM) principle, and is a new set of theory system, which specially aims at machine learning issues under the circumstances of small-sample. Support vector machine (SVM) based on the SLT is a new approach and research field in machine learning because of its advantages such as firm mathematic theory foundation, strict theory analysis, complete theory, global optimization as well as good adaptability and generalization. SVM improves the algorithm generalization effectively and minimizes the empirical risk simultaneously. It has good latent application values and development prospects.First of all, we worked over the solution method of SVM and discussed several typical methods improved performance of SVM. They are quadratic programming, decomposition algorithm, increment algorithm and integrated algorithm of several advanced methods. The differences of capability between SVM and Neural Network (NN) are reflected when NN is applied to the same sample. In this paper, we also elaborated on the characteristic and advantage of SVM.The performance of SVM is decided on the parameters of kernel function and the error punishing parameter. In order to get optimal parameters automatically, a new approach based on genetic algorithm (GA) was proposed, which can acquire the best parameters of SVM. The experiment result shows that the GA-SVM is feasible and effectual.In this paper, two novel algorithms are proposed. It's too much time cost when classical support vector machine handling with large datasets. A FCMSVM (SVM based on fuzzy c-means clustering) is proposed. A distributed fuzzy support vector machine is constructed. Simulation result shows that our method speeds up training while possesses high precision of SVM and can meet actual demands. Another improvement method based on distance difference was proposed. The result of experiment demonstrate that the algorithm can get the SVM with better recognition.In the end, two examples, gear box and slowing box, are used to make a study of the applicable of SVM in fault diagnosis.