A least-squares approximation of a smooth function of means is defined, by projecting the smooth function on a random span of functions of the same analytic form. Asymptotics of this approximation is studied. The variance of a smooth function of means can be lowered by the variance of its least-squares approximation, for all finite sample sizes. Finally, an efficient estimator of the asymptotic variance of a smooth function of means is defined and studied.
Least-squares approximation of smooth functions of means
PALLINI, ANDREA
2004-01-01
Abstract
A least-squares approximation of a smooth function of means is defined, by projecting the smooth function on a random span of functions of the same analytic form. Asymptotics of this approximation is studied. The variance of a smooth function of means can be lowered by the variance of its least-squares approximation, for all finite sample sizes. Finally, an efficient estimator of the asymptotic variance of a smooth function of means is defined and studied.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
