The law of the iterated logarithm for the integrated squared deviation of a kernel density estimator
Giné, Evarist ; Mason, David M.
Bernoulli, Tome 10 (2004) no. 2, p. 721-752 / Harvested from Project Euclid
Let $f_{n,K}$ denote a kernel estimator of a density f in $\R$ such that $\int_\R f ^p(x)\d x \infty$ for some p>2. It is shown, under quite general conditions on the kernel K and on the window sizes, that the centred integrated squared deviation of $f_{n,K}$ from its mean, $\|f_{n,K}-\E f_{n,K}\|_2^2-\E\|f_{n,K}-\E f_{n,K}\|_2^2$ satisfies a law of the iterated logarithm (LIL). This is then used to obtain an LIL for the deviation from the true density, $\|f_{n,K}-f\|_2^2-\E\|f_{n,K}-f\|_2^2$ . The main tools are the Komlós-Major-Tusnády approximation, a moderate-deviation result for triangular arrays of weighted chi-square variables adapted from work by Pinsky, and an exponential inequality due to Giné, Latała and Zinn for degenerate U-statistics applied in combination with decoupling and maximal inequalities.
Publié le : 2004-08-14
Classification:  integrated squared deviation,  kernel density estimator,  law of the iterated logarithm
@article{1093265638,
     author = {Gin\'e, Evarist and Mason, David M.},
     title = {The law of the iterated logarithm for the integrated squared deviation of a kernel density estimator},
     journal = {Bernoulli},
     volume = {10},
     number = {2},
     year = {2004},
     pages = { 721-752},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1093265638}
}
Giné, Evarist; Mason, David M. The law of the iterated logarithm for the integrated squared deviation of a kernel density estimator. Bernoulli, Tome 10 (2004) no. 2, pp.  721-752. http://gdmltest.u-ga.fr/item/1093265638/