A Robust Subset Selection Procedure for Location Parameter Case Based on Hodges-Lehmann Estimators
Abstract
This paper deals with a robust subset selection procedure based on Hodges-Lehmann estimators of location parameters. An improved formula for the estimated standard error of Hodges-Lehmann estimators is considered. Also, the degrees of freedom of the studentized Hodges-Lehmann estimators are investigated and it is suggested to use 0.8n instead of n - 1. The proposed procedure is compared with the other subset selection procedures and its is shown to have good efficiency for heavy-tailed distributions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1988
- Accession Number
- ADA204289
Entities
People
- Kang S. Lee
Organizations
- Purdue University