Exact Upper and Lower Bounds on the Misclassification Probability
Iosif F Pinelis · IEEE Transactions on Information Theory · 2019
Exact upper and lower bounds on the best possible misclassification probability for a finite number of classes are obtained in terms of the total variation norms of the differences between the sub-distributions over the classes. These bounds are compared with the exact upper and lower bounds in terms of the conditional entropy obtained by Feder and Merhav.