Home > Research > Publications & Outputs > Comparing two samples by penalized logistic reg...

Links

Text available via DOI:

View graph of relations

Comparing two samples by penalized logistic regression

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published
<mark>Journal publication date</mark>2008
<mark>Journal</mark>Electronic Journal of Statistics
Volume2
Number of pages17
Pages (from-to)564-580
Publication StatusPublished
<mark>Original language</mark>English

Abstract

Inference based on the penalized density ratio model is proposed and studied. The model under consideration is specified by assuming that the log–likelihood function of two unknown densities is of some parametric form. The model has been extended to cover multiple samples problems while its theoretical properties have been investigated using large sample theory. A main application of the density ratio model is testing whether two, or more, distributions are equal. We extend these results by arguing that the penalized maximum empirical likelihood estimator has less mean square error than that of the ordinary maximum likelihood estimator, especially for small samples. In fact, penalization resolves any existence problems of estimators and a modified Wald type test statistic can be employed for testing equality of the two distributions. A limited simulation study supports further the theory.