On a Conditional Procedure for Selecting a Subset Containing the Best of Several Binomial Populations.
Abstract
Let (Pi sub 1), (Pi sub 2),..., (Pi sub k) be k independent binomial populations such that the observation (X sub i) from (pi sub i) has frequency b(n; p sub 1) for i=1,...,k. The authors consider the problem of selecting a subset of these k populations which contains the population associated with the largest p sub i. The authors derive a lower bound for the infimum of the probability of a correct selection by using Gupta-Huang procedure which is based on the statistic max (X sub j) - (X sub i) being conditioned on the total number of observations, the summation from 1 to k of (X sub i). An upper bound is discussed for the expected size of the selected subset for some special cases. Tables are computed for the conservative values of the constant which is needed to carry out the procedure, for k=2(1)10, n=1(1)10 and P*=0.75, 0.90, 0.95 and 0.99. (Modified author abstract)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1973
- Accession Number
- AD0769239
Entities
People
- Deng Yuang Huang
- Shanti Gupta
- Wen-tao Huang
Organizations
- Purdue University