Detection of Real Estate Mortgage Fraud in a Commercial Banks with Binary Response Model

Abstrcat Fraudulent behaviors became a problem that concerns the bank industries. This article demonstrates the performance of binary response models for fraud detection and implements models for misclassification in the response variable. A database from the commercial bank that contains honest and fraudulent claims is used. The estimation of the probability of omission provides an estimate of the percentage of fraudulent claim that are not detected by the logistic regression model.