Confuser rejection modeling

In this paper we present a method of estimating the performance of a classifier system based on a reported confusion matrix. We develop a method of fitting parameters to observed confusion matrices, such that the expected performance of non-forced decision detection versus the probability of not-in-library reports can be estimated based on the forced decision confusion matrix. The approach also lends itself to a general methodology for modeling classes of confusers in a statistical manner, and can in principle also be extended to modeling clutter severity.