Using Stopping Rules to Bound the Mean Integrated Squared Error in Density Estimation
Martinsek, Adam T.
Ann. Statist., Tome 20 (1992) no. 1, p. 797-806 / Harvested from Project Euclid
Suppose $X_1,X_2,\ldots,X_n$ are i.i.d. with unknown density $f$. There is a well-known expression for the asymptotic mean integrated squared error (MISE) in estimating $f$ by a kernel estimate $\hat{f}_n$, under certain conditions on $f$, the kernel and the bandwidth. Suppose that one would like to choose a sample size so that the MISE is smaller than some preassigned positive number $w$. Based on the asymptotic expression for the MISE, one can identify an appropriate sample size to solve this problem. However, the appropriate sample size depends on a functional of the density that typically is unknown. In this paper, a stopping rule is proposed for the purpose of bounding the MISE, and this rule is shown to be asymptotically efficient in a certain sense as $w$ approaches zero. These results are obtained for data-driven bandwidths that are asymptotically optimal as $n$ goes to infinity.
Publié le : 1992-06-14
Classification:  Density estimation,  stopping rule,  sequential estimation,  asymptotic efficiency,  mean integrated squared error,  62G07,  62L12,  62G20
@article{1176348657,
     author = {Martinsek, Adam T.},
     title = {Using Stopping Rules to Bound the Mean Integrated Squared Error in Density Estimation},
     journal = {Ann. Statist.},
     volume = {20},
     number = {1},
     year = {1992},
     pages = { 797-806},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176348657}
}
Martinsek, Adam T. Using Stopping Rules to Bound the Mean Integrated Squared Error in Density Estimation. Ann. Statist., Tome 20 (1992) no. 1, pp.  797-806. http://gdmltest.u-ga.fr/item/1176348657/