Bayes Risk for the Test of Location-The Infinite Dimensional Case.
PURDUE UNIV LAFAYETTE IND DEPT OF STATISTICS
Pagination or Media Count:
Let X1, X2 be independent normal with mean vector theta and covariance matrix I. Let the null hypothesis be normal with zero mean and covariance matrix sigma. Quadratic loss theta primed A theta and constant loss are considered. Rubin and Sethuraman obtained asymptotic results for the above test in the case of a finite dimensional parameter space. New asymptotic results have been obtained for the case in which the parameter space is infinite dimensional. This development is motivated by the need to extend the Bayes Risk Efficiency analysis to time series problems and to problems in which the alternative hypothesis is a function space. Author
- Statistics and Probability