Research Of Multi-Class Classification Support Vector Machine

Support Vector Machine is a new machine learning algorithm proposed by Vapnik and his group at AT&T Bell laboratory for classification and regression question. It based on structural risk minimization can effectively solve the over study problem and has the good extension and the better classified accuracy. It has become the new research hotspot after the research of pattern recognition and artificial nerve net and will push the development in machine learning theory and technology.Recently, Support Vector Machine is well applied in pattern recognition, function approxmate,data mining and text auto categorization. Traditional Support Vector Machine is developed for binary classification problems, while for practical problems such as data mining and text categorization that need to deal with huge and multi-category data. It is one of the important challenges how to solve large scale and multi-class problemes in recent years.Statistical learning theory is introduced in this paper and Support Vector Machine based on this theory is researched deeply. Training algorithm, classifying algorithm and kernel function are discussed. The present research and application of Support Vector Marchine are expatiated and point out some important issues which are to be resloved when researchers do further research of Support Vector Machine. The multi-class classification methods are summarized including one-agaist-rest, one-agaist-one and decision directed acyclic graph support vector machine, and their advantage, disadvantage and capability are compared.Finally, the disadvantages of the existing methods of Support Vector Machine multi-class classification are analyzed and compared in this paper, such as 1-a-r is difficult to train and the classifying speed of 1-a-1 is slow. To solve these problems, a parallel multi-class Support Vector Machine based on...