On asymptotic minimaxity of kernel-based tests

Ermakov, Michael

ESAIM: Probability and Statistics, Tome 7 (2003), pp. 279-312

Résumé

In the problem of signal detection in gaussian white noise we show asymptotic minimaxity of kernel-based tests. The test statistics equal $L_2$-norms of kernel estimates. The sets of alternatives are essentially nonparametric and are defined as the sets of all signals such that the $L_2$-norms of signal smoothed by the kernels exceed some constants ${\rho }_{ϵ}>0$. The constant ${\rho }_{ϵ}$ depends on the power $ϵ$ of noise and ${\rho }_{ϵ}\to 0$ as $ϵ\to 0$. Similar statements are proved also if an additional information on a signal smoothness is given. By theorems on asymptotic equivalence of statistical experiments these results are extended to the problems of testing nonparametric hypotheses on density and regression. The exact asymptotically minimax lower bounds of type II error probabilities are pointed out for all these settings. Similar results are also obtained for the problems of testing parametric hypotheses versus nonparametric sets of alternatives.

Détail

DOI : 10.1051/ps:2003013

Classification : 62G10, 62G20
Keywords: nonparametric hypothesis testing, kernel-based tests, goodness-of-fit tests, efficiency, asymptotic minimaxity, kernel estimator