期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2017
卷号：17
期号：12
页码：145-150
出版社：International Journal of Computer Science and Network Security
摘要：In this article, we are going to study the linear support vectors and their performance in the related classification issues. Using the linear support vectors (SVM's) in the classification issues is a new approach that in recent years is considered by many scientists. It was used in a wide range of applications including OCR, Handwriting recognition, guidance signs diagnosis and etc. SVM approach is in a way that in the training phase, it is tried to choose the limit of decision-making (Decision Boundary) is such a way that its minimum distance to each of the considered categories stays maximum. This kind of choice helps our decision in practice to tolerate the noisy condition very well and has a good response. This way of selecting the boundary is based on the points that are named as support vectors. At first we study the concepts such as generalization of a pattern recognition machine and then the VC dimension that has a great application in the concept of classification machines. And then we describe the linear and non-linear support vectors and Kernel functions. And eventually, we will study the VC dimension for some of these functions.
关键词：Support Vector Machine; VC Dimension; Mercer; Linear SVM; Nonlinear SVM RBF Kernel; Medical Data
