Testing convex hypotheses on the mean of a Gaussian vector. Application to testing qualitative hypotheses on a regression function
Baraud, Yannick ; Huet, Sylvie ; Laurent, Béatrice
Ann. Statist., Tome 33 (2005) no. 1, p. 214-257 / Harvested from Project Euclid
In this paper we propose a general methodology, based on multiple testing, for testing that the mean of a Gaussian vector in ℝn belongs to a convex set. We show that the test achieves its nominal level, and characterize a class of vectors over which the tests achieve a prescribed power. In the functional regression model this general methodology is applied to test some qualitative hypotheses on the regression function. For example, we test that the regression function is positive, increasing, convex, or more generally, satisfies a differential inequality. Uniform separation rates over classes of smooth functions are established and a comparison with other results in the literature is provided. A simulation study evaluates some of the procedures for testing monotonicity.
Publié le : 2005-02-14
Classification:  Tests of qualitative hypotheses,  nonparametric test,  test of positivity,  test of monotonicity,  test of convexity,  rate of testing,  Gaussian regression,  62G10,  62G20
@article{1112967705,
     author = {Baraud, Yannick and Huet, Sylvie and Laurent, B\'eatrice},
     title = {Testing convex hypotheses on the mean of a Gaussian vector. Application to testing qualitative hypotheses on a regression function},
     journal = {Ann. Statist.},
     volume = {33},
     number = {1},
     year = {2005},
     pages = { 214-257},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1112967705}
}
Baraud, Yannick; Huet, Sylvie; Laurent, Béatrice. Testing convex hypotheses on the mean of a Gaussian vector. Application to testing qualitative hypotheses on a regression function. Ann. Statist., Tome 33 (2005) no. 1, pp.  214-257. http://gdmltest.u-ga.fr/item/1112967705/