| PRTools contents |
SETCOST
W = SETCOST(W,COST,LABLIST)
The classification cost matrix of the dataset W is reset to COST. W has to be a trained classifier. COST should have size [C,C+1], if C is the number of classes assigned by W. COST(I,J) are the costs of classifying an object of class I as class J. Column C+1 generates an alternative reject class and may be omitted, yielding a size of [C,C]. An empty cost matrix, COST = [] (default) is interpreted as COST = ONES(C) - EYE(C) (identical costs of misclassification).
In LABLIST the corresponding class labels may be supplied. LABLIST may have only class names of the existing labels assigned by W, stored in W.LABELS.
| PRTools contents |