Support vector machine (SVM) is the most popular classifier based on a linear discriminant function. It is ideally suited for binary classification. It has been studied extensively in several pattern recognition applications and in data mining. It has become a baseline standard for classification because of excellent software packages that have been developed systematically over the past three decades. In this chapter, we introduce SVM-based classification and some of the essential properties related to classification. Specifically we deal with linear SVM that is ideally suited to deal with linearly separable classes.

Social and information networks are playing an important role in several applications. One of important problems here is classification of entities in the networks. In this chapter, we discuss several notions associated with social networks and the role of linear classifiers.

Kernel Support Vector Machine (SVM) is useful to deal with nonlinear classification based on a linear discriminant function in a high-dimensional (kernel) space. Linear SVM is popularly used in applications involving high-dimensional spaces. However, in low-dimensional spaces, kernel SVM is a popular nonlinear classifier. It employs kernel trick which permits us to work in the input space instead of dealing with a potentially high-dimensional, even theoretically infinite dimensional, kernel (feature) space. Also kernel trick has become so popular that it is used in a variety of other pattern recognition and machine learning algorithms.

This work reviews the state of the art in SVM and perceptron classifiers. A Support Vector Machine (SVM) is easily the most popular tool for dealing with a variety of machine-learning tasks, including

This work reviews the state of the art in SVM and perceptron classifiers. A Support Vector Machine (SVM) is easily the most popular tool for dealing with a variety of machine-learning tasks, including classification. SVMs are associated with maximizing the margin between two classes. The concerned optimization problem is a convex optimization guaranteeing a globally optimal solution. The weight vector associated with SVM is obtained by a linear combination of some of the boundary and noisy vectors. Further, when the data are not linearly separable, tuning the coefficient of the regularization term becomes crucial. Even though SVMs have popularized the kernel trick, in most of the practical applications that are high-dimensional, linear SVMs are popularly used. The text examines applications to social and information networks. The work also discusses another popular linear classifier, the perceptron, and compares its performance with that of the SVM in different application areas.

Linear discriminant functions (LDFs) have been successfully used in pattern classification. Perceptrons and Support Vector Machines (SVMs) are two well-known members of the category of linear discriminant functions that have been popularly used in classification. In this chapter, we introduce the notion of linear discriminant function and some of the important properties associated with it.