Extended Optimality of Sequential Probability Ratio Tests
Irle, Albrecht
Ann. Statist., Tome 12 (1984) no. 1, p. 380-386 / Harvested from Project Euclid
The problem of sequentially testing two simple hypotheses for a stochastic process is considered. It is shown that, for arbitrary distributions $P_0$ and $P_1$, the following optimality holds for an SPRT which stops on its boundaries: If $\alpha$ and $\beta$ represent the error probabilities of the SPRT and a competing test has error probabilities $\alpha' \leq \alpha$ and $\beta' \leq \beta$ then $E_0g(D_{\tau'}) \geq E_0g(D_\tau)$ for any convex function $g$ satisfying some minor requirement, provided $P_1(\tau' < \infty) = 1$ for the competing test. Here $D_\tau$ and $D_{\tau'}$ denote the terminal likelihood ratios under the SPRT and the competitor. An analogous statement holds for expectation under $P_1$, and several applications of this optimality result are given.
Publié le : 1984-03-14
Classification:  Sequential probability ratio test,  optimality,  admissibility,  62L10,  62L15
@article{1176346416,
     author = {Irle, Albrecht},
     title = {Extended Optimality of Sequential Probability Ratio Tests},
     journal = {Ann. Statist.},
     volume = {12},
     number = {1},
     year = {1984},
     pages = { 380-386},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176346416}
}
Irle, Albrecht. Extended Optimality of Sequential Probability Ratio Tests. Ann. Statist., Tome 12 (1984) no. 1, pp.  380-386. http://gdmltest.u-ga.fr/item/1176346416/