A Sharp Necessary and Sufficient Condition for Inadmissibility of Estimators in a Control Problem
Srinivasan, C.
Ann. Statist., Tome 12 (1984) no. 1, p. 927-944 / Harvested from Project Euclid
Suppose $\mathbf{x} = (x_1, \cdots, x_m)^t$ is an $m$-variate normal random variable with mean vector $\mathbf{\theta} = (\theta_1, \cdots, \theta_m)^t$ and identity dispersion matrix. We consider the control problem which, in canonical form, is the problem of estimating $\mathbf{\theta}$ with respect to the loss $L(\theta, \delta) = (1 - \theta^t\delta)^2,$ where $\delta(x) = (\delta_1(x), \cdots, \delta_m(x))^t$. A necessary and sufficient condition for the admissibility of spherically symmetric generalized Bayes $\delta(x)$ is given in terms of a Dirichlet problem. This condition is also equivalent to recurrence of a diffusion process and insolubility of certain elliptic boundary value problems.
Publié le : 1984-09-14
Classification:  Inadmissibility,  formal Bayes estimators,  control problem,  62C15,  62F10,  62P20
@article{1176346712,
     author = {Srinivasan, C.},
     title = {A Sharp Necessary and Sufficient Condition for Inadmissibility of Estimators in a Control Problem},
     journal = {Ann. Statist.},
     volume = {12},
     number = {1},
     year = {1984},
     pages = { 927-944},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176346712}
}
Srinivasan, C. A Sharp Necessary and Sufficient Condition for Inadmissibility of Estimators in a Control Problem. Ann. Statist., Tome 12 (1984) no. 1, pp.  927-944. http://gdmltest.u-ga.fr/item/1176346712/