Optimal rates for plug-in estimators of density level sets
Rigollet, Philippe ; Vert, Régis
Bernoulli, Tome 15 (2009) no. 1, p. 1154-1178 / Harvested from Project Euclid
In the context of density level set estimation, we study the convergence of general plug-in methods under two main assumptions on the density for a given level λ. More precisely, it is assumed that the density (i) is smooth in a neighborhood of λ and (ii) has γ-exponent at level λ. Condition (i) ensures that the density can be estimated at a standard nonparametric rate and condition (ii) is similar to Tsybakov’s margin assumption which is stated for the classification framework. Under these assumptions, we derive optimal rates of convergence for plug-in estimators. Explicit convergence rates are given for plug-in estimators based on kernel density estimators when the underlying measure is the Lebesgue measure. Lower bounds proving optimality of the rates in a minimax sense when the density is Hölder smooth are also provided.
Publié le : 2009-11-15
Classification:  density level sets,  kernel density estimators,  minimax lower bounds,  plug-in estimators,  rates of convergence
@article{1262962230,
     author = {Rigollet, Philippe and Vert, R\'egis},
     title = {Optimal rates for plug-in estimators of density level sets},
     journal = {Bernoulli},
     volume = {15},
     number = {1},
     year = {2009},
     pages = { 1154-1178},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1262962230}
}
Rigollet, Philippe; Vert, Régis. Optimal rates for plug-in estimators of density level sets. Bernoulli, Tome 15 (2009) no. 1, pp.  1154-1178. http://gdmltest.u-ga.fr/item/1262962230/