In this dissertation, some problems of support vector machine algorithms are analyzed. To solve the problem of uneven size of classes and other problems, two new support vector machine algorithms and two new classification methods are showed.
The main research in this paper can be classed as follows:
(1) An overview on a variety of classification algorithms for support vector machine (SVM) is given. Some algorithms such as v-SVM, One-class SVM, RSVM, weighted SVM and LS-SV...