ON SOME CLASSES OF SELECTION PROCEDURES BASED ON RANKS.
Abstract
The paper deals with some nonrandomized ranking and selection procedures based on ranks using the subset selection approach. The main problem is to select a subset of k given populations, which contains the 'best' population with probability at least P*. The random variables associated with a fixed population are assumed to be independent identically distributed with a continuous distribution function depending on a single parameter. This parameter is assumed to stochastically order the k distribution functions, and the 'best' population is the stochastically largest (smallest) population. The procedures presented depend on the individual observations of a given populations only through their ranks in the combined sample. In some preference-type tests or lost-data problems, these ranks may be the only information available to an experimenter. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1969
- Accession Number
- AD0689873
Entities
People
- Gary C. Mcdonald
- Shanti Gupta
Organizations
- Purdue University