Author : Rashmi Pathak and Shyam Gupta

Machine Learning means “learning through Machines”. We make machine learn and predict the behavior in order to find a solution to a problem. Machine is set to make various predictions based on the learning mechanisms that have been incorporated in them. There are various techniques through which machines can learn. Learning can be supervised, Unsupervised or Semi-supervised. Under these learning schemes we have various classifications. Under Supervised learning we have Classification and Regression. Classification works on continuous values and Regression works on discrete values. Support Vector Machine is an efficient classifier which are mostly sort of linear and comes under supervised method of learning. SVM also find its application in real life for Face Detection, Bioinformatics, Handwriting recognition, image classification and many others. Though this paper we want to review the support vector machines concept and represent it in primal form. We have also stated dual formulation here as can be used to introduce the feature mapping that is mostly used in nonlinearly cases those are separable

Keywords: SVM, Classifier, Dual formation, Machine learning, Artificial Intelligence, Feature selection, Embedded methods, Support Vector machines, Mathematical programming