Supervised and Unsupervised Neural Networks –Topology, Learning and Applications

Abstract

Modern digital computers outperform humans in the domain of numeric computation and related symbol manipulation. However, humans can effortlessly solve complex perceptual problems (like recognizing a man in a crowd from a mere glimpse of his face) at such a high speed and extent as to dwarf the world’s fastest computer. Support Vector Machines (SVM) classifier are a set of related supervised learning methods, for which solving classification and regression tasks is formulated as quadratic programming (QP) problems. SVMs may be defined as a classification method that determines that maximum-margin hyperplane. In the case of basic linear classification, a SVM creates a maximum margin hyperplane that lies in a transformed input space. Given binary choice training examples (labeled either `yes' or `no'), a maximum-margin hyperplane divides the `yes' and `no' examples, such that the distance from the closest examples, i.e. the margin, to the hyperplane is maximized.