International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Support Vector Machines (SVMs) are nothing but machines that build support vectors for classification process. SVMs can formulate both linear and non-linear decision boundaries with good generalization ability and they are based on Statistical Learning Theory (SLT). Generally classification using SVMs is same as solving an optimization problem because strength of SVMs lies in solving optimization techniques. At present SVMs have become excellent areas for research which are also powerful tools for most of the machine learning tasks. These optimization problems deal with not only convex problems but also non convex such as semi infinite programming, bi-level programming and integer programming.