POSTERIOR INFERENCE ON THE DEGREES OF FREEDOM PARAMETER IN MULTIVARIATE-T REGRESSION-MODELS

被引：17

作者：

CHIB, S

OSIEWALSKI, J

STEEL, MFJ

机构：

[1] ACAD ECON KRAKOW,KRAKOW,POLAND

[2] TILBURG UNIV,5000 LE TILBURG,NETHERLANDS

来源：

ECONOMICS LETTERS | 1991年 / 37卷 / 04期

关键词：

D O I：

10.1016/0165-1765(91)90076-W

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

This paper considers the nonlinear regression model with errors that follow the multivariate Student-t distribution with nu-degrees of freedom. We specify general conditions on the overall prior structure under which the prior of nu is not updated by the sample information, and provide an example in which learning about nu is not precluded.

引用

页码：391 / 397

页数：7

共 7 条

[1] BAYES PREDICTION IN REGRESSIONS WITH ELLIPTICAL ERRORS [J].