Discriminant Analysis When the Initial Samples are Misclassified II: Non-Random Misclassification Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Learning kernel logistic regression in the presence of class label noise (2014, background): “Part of the reason for this may be that uniform/symmetric label noise is relatively harmless [21,22,12,27].”
Classification in the Presence of Label Noise: A Survey (2014, methodology): “Answering this question could lead to more complex and realistic models of label noise in the line of e.g. [5], [56], [67], [70]–[72], [90], [91], [122], [227]–[230], [235], [251].”
Statistical fraud detection: A review (2002, background): “Some work has addressed misclassification of training samples (e.g., Lachenbruch, 1966, 1974; Chhikara and McKeon, 1984), but not in the context of fraud detection as far as we are aware.”
Linear discriminant analysis with misallocation in training samples (1984, methodology): “If we assume u=1 and a l ua 2=0, then one obtains the complete separation model of lachenbruch (1974).”