TESTAUC

Multiclass error area under the ROC

    E = TESTAUC(A*W)
    E = TESTAUC(A,W)
    E = A*W*TESTAUC

Input
A	Dataset to be classified
W	Classifier

Output
E	Error, Area under the ROC

Description

The area under the ROC is computed for the datset A w.r.t. the  classifer W. The estimator is based on a rank analysis of the classifier  outcomes. Ties are broken by a two-way sorting and averaging.

The multiclass situation is solved by averaging over all outcomes of  the one-against-rest ROCs.

Note that E is an error and not a performance measure like the AUC often  used in literature.

See also

datasets, mappings, testc, roc,