The method organizes transaction sums into an optimal number of clusters and define deviations within each cluster separately. Therefore, this method performes data clustering in order to determine the best arrangement of values into different classes. The method is seeking to minimize each class’s average deviation from the class mean and at the same time it is maximizing each class’s deviation from the means of the other groups. So method cluster data in a manner that it reduces the variance within classes and maximize the variance between classes.