On Some Gamma-Minimax Subset Selection and Multiple Comparison Procedures.
PURDUE UNIV LAFAYETTE IND DEPT OF STATISTICS
Pagination or Media Count:
The use of partial or incomplete prior information in statistical inference has led to the development of the gamma-minimax criterion which allows one to select a decision rule that minimizes the maximum expected risk over gamma. In this paper, the authors are concerned with the problem of selecting a subset containing the best population and containing all superior populations and of multiple comparison procedures which are optimal by using gamma-minimax criterion. Some applications are discussed. Asymptotic optimal nonparametric procedures are also considered. Author
- Statistics and Probability