Stochastic Reduction of Loss in Estimating Normal Means by Isotonic Regression
Kelly, Robert E.
Ann. Statist., Tome 17 (1989) no. 1, p. 937-940 / Harvested from Project Euclid
Consider the problem of estimating the ordered means $\mu_1 \leq \mu_2 \leq \cdots \leq \mu_k$ of independent normal random variables, $Y_1, Y_2, \cdots, Y_k$. It is shown that the absolute error of each component $\hat{\mu}_i$ of the isotonic regression estimator is stochastically smaller than that of the usual estimator $Y_i$. Thus $\hat{\mu}_i$ is superior to $Y_i$ under any nonconstant loss which is a nondecreasing function of absolute error.
Publié le : 1989-06-14
Classification:  Isotonic regression,  loss function,  maximum likelihood,  order restriction,  stochastic ordering,  62F10,  62E99,  62C99,  60E15
@article{1176347153,
     author = {Kelly, Robert E.},
     title = {Stochastic Reduction of Loss in Estimating Normal Means by Isotonic Regression},
     journal = {Ann. Statist.},
     volume = {17},
     number = {1},
     year = {1989},
     pages = { 937-940},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176347153}
}
Kelly, Robert E. Stochastic Reduction of Loss in Estimating Normal Means by Isotonic Regression. Ann. Statist., Tome 17 (1989) no. 1, pp.  937-940. http://gdmltest.u-ga.fr/item/1176347153/