Statistical Selection Procedures in Multivariate Models,
Abstract
Selection and ranking problems have been studied over the last thirty years, generally under one of two formulations: Bechhofer's indifference zone approach and Gupta's subset selection approach. This paper deals with subset selection. Subset selection procedures in multivariate models are briefly reviewed. These include: (1) Procedures for selecting the best component in a multivariate normal population in terms of the component means as well as the component variances; (2) Procedures for selecting the best from several multivariate normal populations in terms of the Mahalanobis distance, the generalized variance, and the multiple correlation coefficient; (3) Procedures (fixed sample size as well as inverse sampling) for selecting the most (least) probable cell in a multinominal distribution; (4) Procedures for selecting the best from several multinomial populations in terms of the Shannon entropy function; and (5) Procedures for choosing the best subset of the predictor variables in a linear regression model. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1986
- Accession Number
- ADA176043
Entities
People
- S. Panchapakesan
- Shanti Gupta
Organizations
- Purdue University