Support Vector Machine Using Multiple Hyperplanes For Rare Class

Support vector machine is a very popular method to classify data from different classes. But it doesn't work well when the number of data in one of the class is very rare. In this article, a new method called Support Vector Machine using Multiple Hy-perplanes will be introduced. Instead of using one hyperplane to classify two classes of data, SVM_MH uses multiple hyperplanes. All the data in the rare class have to on the side of the hyperplanes that parallel to the normal vectors. The classifier of the new method is more strict than that of SVM. SVM is also a SVM_MH method with only one hyperplane. Kernels can be used in the new method.