The concept of SVM (Support Vector Machine) was
introduced by Vapnik and co-workers [4]. It gains
popularity because it offers the attractive features and
powerful machinery to tacklethe problem of classification
i.e., we need to know which belongs to which group and
promising empirical performance.
The SVM is based on statistical learning theory. SVM’s
better generalization performance is based on the principle
of Structural Risk Minimization (SRM) [4]. The concept of
SRM is to maximize the margin of class separation. The
SVM was defined for two-class problem and it looked for
optimal hyper-plane, which maximized the distance, the
margin, between the nearest examples of both classes,
named SVM [5]. For information detail about SVM can be
seen in [4,5,6].
We have utilised radial basis function for its kernel
function. The input feature sets were the directional features
(169-dimension). All the SVM’s are trained with the
respective training feature sets and the results explored by
using separate test data which are lowercase and uppercase
letters