Uniformly Minimum Variance Unbiased Estimation for Distributions with Support Dependign on Two Parameters

  • Hong, Chong-Sun (Department of Statistics, Sung Kyun Kwan University, 3-53 Nyungryun Dong, Chongro-ku, Seoul 110-745) ;
  • Park, Hyun-Jip (Department of Statistics, Sung Kyun Kwan University, 3-53 Nyungryun Dong, Chongro-ku, Seoul 110-745) ;
  • Lee, Chong-Cheol (Department of Statistics, Sung Kyun Kwan University, 3-53 Nyungryun Dong, Chongro-ku, Seoul 110-745)
  • 발행 : 1995.06.01

초록

When a random sample is taken from a certain class of discrete and continuous distributions whose support depend on two parameters, we could find that there exists the complete and sufficient statistic for parameters which belong to a certain class, and fomulate the uniformly minimum variance unbiased estimator (UMVUE) of any estimable function. Some UMVUE's of parametric functions are illustrated for the class of the distribution. Especially, we find that the UMVUE of some estimable parametric function from the truncated normal distribution could be expressed by the version of the Mill's ratio.

키워드

참고문헌

  1. Sankhya v.18 On Unbiased Estimates of Uniformly Minimum Variance Bahadur,R.R.
  2. Order Statistics David,H.Q.
  3. The American Staistician v.43 no.1 Uniformly Minimum Variance Unbiased Estimation for Discrete Distributions With Support Depending on the Parameters Ghosh, Malay;Gauri Sankar Datta
  4. The Advanced Theory of Statistics(3rd) v.1;2 Kendall,M.;Stuart,A.
  5. Theory of Point Estimation Lehmann,E.L.
  6. The American Statistician v.26 no.1 Completeness and Unbiased Estimator Stigler, Stephen M.
  7. Annals of Mathematical Statistics v.30 Unbiased Estimation:Functions of Location and Scale Parameters Tate,R.F.
  8. The Theory of Statistical Inference Zacks, Shelemyahu