Kernels Based Tests with Non-asymptotic Bootstrap Approaches for Two-sample Problems

Considering either two independent i.i.d. samples, or two independent samples generated from a heteroscedastic regression model, or two independent Poisson processes, we address the question of testing equality of their respective distributions. We first propose single testing procedures based on a general symmetric kernel. The corresponding critical values are chosen from a wild or permutation bootstrap approach, and the obtained tests are exactly (and not just asymptotically) of level . We then introduce an aggregation method, which enables to overcome the difficulty of choosing a kernel and/or the parameters of the kernel. We derive non-asymptotic properties for the aggregated tests, proving that they may be optimal in a classical statistical sense.

[1]  Jeffrey D. Hart,et al.  Testing the equality of two regression curves using linear smoothers , 1991 .

[2]  Alexander J. Smola,et al.  Learning with kernels , 1998 .

[3]  John G. Saw,et al.  On comparing two poisson intensity functions , 1980 .

[4]  Bernhard Schölkopf,et al.  A Kernel Method for the Two-Sample-Problem , 2006, NIPS.

[5]  M. Fromonta,et al.  Adaptive tests of homogeneity for a Poisson process , 2010 .

[6]  Zaïd Harchaoui,et al.  A Fast, Consistent Kernel Two-Sample Test , 2009, NIPS.

[7]  Y. Baraud.  Non-asymptotic minimax rates of testing in signal detection , 2002 .

[8]  P. Bickel.  A Distribution Free Version of the Smirnov Two Sample Test in the $p$-Variate Case , 1969 .

[9]  E. Giné,et al.  On the Bootstrap of $U$ and $V$ Statistics , 1992 .

[10]  Enno Mammen,et al.  Bootstrap, wild bootstrap, and asymptotic normality , 1992 .

[11]  Michael Tiefelsdorf.  The regression model , 2000 .

[12]  B. Laurent,et al.  ADAPTIVE TESTS OF LINEAR HYPOTHESES BY MODEL SELECTION , 2003 .

[13]  Bernhard Schölkopf,et al.  Characteristic Kernels on Groups and Semigroups , 2008, NIPS.

[14]  Peter Hall,et al.  Bootstrap test for difference between means in nonparametric regression , 1990 .

[15]  Qi Li,et al.  Nonparametric testing the similarity of two unknown density functions: local power and bootstrap analysis , 1999 .

[16]  Daryl J. Daley,et al.  General theory and structure , 2008 .

[17]  S. Halim,et al.  Wild Bootstrap Tests , 2007, IEEE Signal Processing Magazine.

[18]  Jayant V. Deshpande,et al.  Testing of two sample proportional intensity assumption for non-homogeneous Poisson processes , 1999 .

[19]  Alexandre B. Tsybakov,et al.  Introduction to Nonparametric Estimation , 2008, Springer series in statistics.