A Restricted Subset Selection Procedure for Selecting the Largest Normal Mean Under Heteroscedasticity

• Activity Number: 31 - Statistical Inference of Causality and Structure
• Type: Contributed
• Date/Time: Sunday, August 7, 2022 : 2:00 PM to 3:50 PM
• Sponsor: IMS
• Author(s): Elena M Buzaianu* and Pinyuen Chen and Lifang Hsu
• Companies: University of North Florida and Syracuse University and Lemoyne College
• Keywords: Expected Subset Size; Probability of a Correct Selection; Restricted Subset Selection

Abstract

In statistical ranking and selection, the subset selection approach allows the researcher to screen a group of populations, selecting a subset which could be studied more intensively later. Such subset selection is useful when a large number of populations are present. However, due to limited resources at a secondary stage, the experimenters might want to put restrictions on the number of populations to be included in the selected subset. Selection with a restriction on the size of the selected subset is called a restricted subset selection. We propose and study a restricted size subset selection procedure for selecting the largest normal mean among k populations with unknown and unequal variances.