Error Exponents of Mismatched Likelihood Ratio Testing

Abstract: We study the problem of mismatched likelihood ratio test. We analyze the type-\RNum{1} and \RNum{2} error exponents when the actual distributions generating the observation are different from the distributions used in the test. We derive the worst-case error exponents when the actual distributions generating the data are within a relative entropy ball of the test distributions. In addition, we study the sensitivity of the test for small relative entropy balls.