Abstract

this paper is the following. The input consists of an alphabet A, along with an input string or sequence X = x 1 x 2 . . xm and a set Y k = y k1 y k2 . . y kn(k) of target strings, each representing a population from which X might originate. Since the target populations may be infinite, we assume that Y k is a reliable representative of the k th population. Classification of X as to which population it is derived from will be based on determination of the target population representative Y c closest to the input. In order to measure distance between strings X and Y, the notion of edit distance DE(X,Y) will be used [6,8]. This involves four single character edit operations: deletion, insertion, equal and unequal substitution, along with associated edit costs. A cost of 0 is associated with equal substitution and

Citations