Abstract
Let tn be the Bayesian estimator of the parameter θ constructed from independent observations in the case of infinite information and probability distribution density with some singularities. It is shown that under certain conditions on the behavior of the densities near their singularities, the normalizing factor ϕ(n) ensuring that Δn=ϕ(n) (tn-θ) has a nontriviallimiting distribution for n →∞ is regularly varying in Karamata's sense. The limiting distribution of Δn is determined.
Similar content being viewed by others
Literature cited
I. A. Ibragimov and R. Z. Khas'minskii, “On moments of generalized Bayesian estimators and maximum likelihood estimators,” Teor. Veroyatn. Primen.,18, No. 3 (1973).
I. A. Ibragimov and R. Z. Khas'minskii, “Asymptotic behavior of location parameter estimators for samples with a continuous density with singularities,” J. Sov. Math.,9, No. 1 (1978).
W. Feller, Introduction to Probability Theory and Its Applications, Wiley (1968).
I. A. Ibragimov and R. Z. Khas'minskii, “Asymptotic behavior of estimators for samples with discontinuous density,” Mat. Sb.,87, No. 4 (1972).
Additional information
Translated from Zapiski Nauchnykh Seminarov Leningradskogo Otdeleniya Matematicheskogo Instituta im. V. A. Steklova AN SSSR, Vol. 74, pp. 66–82, 1977.
I would like to acknowledge the guidance of I. A. Ibragimov in the course of this research.
Rights and permissions
About this article
Cite this article
Bakirov, N.K. The likelihood ratio for densities with singularities. J Math Sci 34, 1403–1413 (1986). https://doi.org/10.1007/BF01085008
Issue Date:
DOI: https://doi.org/10.1007/BF01085008