amboS Ve use lowercase bold letters to denote vectors and uppercase italic letters to denote matrices. The following list shows the symbols used in the book variable associated with x A inverse of matrix a transpose of matrix A B set of bounded support vector indices b i bias term of the ith hyperplane margin parameter degree of a polynomial kernel p(x) mapping function from x to the feature space parameter for a radial basis function kernel dimension of the feature space number of training data number of input variables number of classes set of support vector indices U set of unbounded support vector indices Euclidean norm of vector x coefficient vector of the ith hyperpla X set for class i training data number of data in the set X ith m-dimensional training data class label 1 or -l for input xi for pattern classification and a scalar output for function approximationSymbols We use lowercase bold letters to denote vectors and uppercase italic letters to denote matrices. The following list shows the symbols used in the book: αi Lagrange multiplier for xi ξi slack variable associated with xi A−1 inverse of matrix A A transpose of matrix A B set of bounded support vector indices bi bias term of the ith hyperplane C margin parameter d degree of a polynomial kernel φ(x) mapping function from x to the feature space γ parameter for a radial basis function kernel K(x, x ) kernel l dimension of the feature space M number of training data m number of input variables n number of classes S set of support vector indices U set of unbounded support vector indices x Euclidean norm of vector x wi coefficient vector of the ith hyperplane Xi set for class i training data |Xi| number of data in the set Xi xi ith m-dimensional training data yi class label 1 or −1 for input xi for pattern classification and a scalar output for function approximation xix