On global and pointwise adaptive estimation
Efromovich, Sam
Bernoulli, Tome 4 (1998) no. 1, p. 273-282 / Harvested from Project Euclid
Let an estimated function belong to a Lipschitz class of order [math] . Consider a minimax approach where the infimum is taken over all possible estimators and the supremum is taken over the considered class of estimated functions. It is known that, if the order [math] is unknown, then the minimax mean squared (pointwise) error convergence slows down from [math] for the case of the given [math] to [math] . At the same time, the minimax mean integrated squared (global) error convergence is proportional to [math] for the cases of known and unknown [math] . We show that a similar phenomenon holds for analytic functions where the lack of knowledge of the maximal set to which the functioncan be analytically continued leads to the loss of a sharp constant. Surprisingly, for the more general adaptive minimax setting where we consider the union of a range of Lipschitz and a range of analytic functions neither pointwise error convergence nor global error convergence suffers an additional slowing down.
Publié le : 1998-06-14
Classification:  analytic and Lipschitz functions,  efficiency,  mean integrated squared error,  mean squared error
@article{1174937296,
     author = {Efromovich, Sam},
     title = {On global and pointwise adaptive estimation},
     journal = {Bernoulli},
     volume = {4},
     number = {1},
     year = {1998},
     pages = { 273-282},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1174937296}
}
Efromovich, Sam. On global and pointwise adaptive estimation. Bernoulli, Tome 4 (1998) no. 1, pp.  273-282. http://gdmltest.u-ga.fr/item/1174937296/