A Feature Selection and Multi-classification Method Based on Improved Support Vector Machine

Support vector machine is a generalized linear classifier, which uses supervised learning to perform binary classification on the sample data, that is, classify the data samples on the optimal hyperplane to obtain the global optimal solution. This has the advantage of making the system structure more precise and clear. SVM is a binary classifier. When performing multi-class classification tasks, a binary SVM classifier is built for each class, and then a multi-class SVM classifier is constructed by combination. At present, SVM multi-classification method has been concerned and studied by scholars at home and abroad. Feature selection is to select the most statistically significant k feature subsets to represent the characteristics of the whole sample under a certain generalization error. This paper proposes a feature selection and multi-classification method based on improved support vector machine. Finally, the comparative experiments show that the classification method proposed in this paper has higher classification accuracy.