Issue No. 09 - September (1974 vol. 23)
G.T. Toussaint , School of Computer Science, McGill University
In a recent note1 C. C. Babu derives an upper bound on the probability of misclassification for the case when part of the training data is incorrectly labeled. Using this notation the bound is given by
G. Toussaint, "Comments on "The Extraction of Pattern Features from Imperfectly Identified Samples"," in IEEE Transactions on Computers, vol. 23, no. , pp. 988-989, 1974.