On the Risk of Histograms for Estimating Decreasing Densities
Birge, Lucien
Ann. Statist., Tome 15 (1987) no. 1, p. 1013-1022 / Harvested from Project Euclid
Suppose we want to estimate an element $f$ of the space $\Theta$ of all decreasing densities on the interval $\lbrack a; a + L \rbrack$ satisfying $f(a^+) \leq H$ from $n$ independent observations. We prove that a suitable histogram $\hat{f}_n$ with unequal bin widths will achieve the following risk: $\sup_{f \in \Theta} \mathbb{E}_f \big\lbrack \int|\hat{f}_n(x) - f(x)|dx \big\rbrack \leq 1.89(S/n)^{1/3} + 0.20(S/n)^{2/3}$, with $S = \operatorname{Log}(HL + 1)$. If $n \geq 39S$, this is only ten times the lower bound given in Birge (1987). An adaptive procedure is suggested when $a, L, H$ are unknown. It is almost as good as the original one.
Publié le : 1987-09-14
Classification:  Histograms,  minimax risk,  decreasing densities,  62G05,  62C20
@article{1176350489,
     author = {Birge, Lucien},
     title = {On the Risk of Histograms for Estimating Decreasing Densities},
     journal = {Ann. Statist.},
     volume = {15},
     number = {1},
     year = {1987},
     pages = { 1013-1022},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176350489}
}
Birge, Lucien. On the Risk of Histograms for Estimating Decreasing Densities. Ann. Statist., Tome 15 (1987) no. 1, pp.  1013-1022. http://gdmltest.u-ga.fr/item/1176350489/