A characterization of probability densities with expected log likelihood

Ryuei Nishii

Research output: Contribution to journalArticlepeer-review

Abstract

In the class of all densities with constraints E bi(X) = Cj, (i = 1, q), it is we I 1 known that the entropy is maximized by the density belonging to the exponential family specified by b.(x). i = 1.q. Our aim in this note is to generalize the result by using expected log likelihood.

Original languageEnglish
Pages (from-to)2657-2662
Number of pages6
JournalCommunications in Statistics - Theory and Methods
Volume18
Issue number7
DOIs
Publication statusPublished - Jan 1 1989

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Fingerprint Dive into the research topics of 'A characterization of probability densities with expected log likelihood'. Together they form a unique fingerprint.

Cite this