On the Amount of Noise Inherent in Bandwidth Selection for a Kernel Density Estimator
Hall, Peter ; Marron, J. S.
Ann. Statist., Tome 15 (1987) no. 1, p. 163-181 / Harvested from Project Euclid
Text on Project Euclid
PDF
Alt PDF
In the setting of kernel density estimation, data-driven bandwidth, i.e., smoothing parameter, selectors are considered. It is seen that there is a well-defined, and surprisingly restrictive, bound on the rate of convergence of any automatic bandwidth selection method to the optimum. The method of least squares cross-validation achieves this bound.
Publié le : 1987-03-14
Classification:  Bandwidth,  cross-validation,  data-driven estimate,  noise,  smoothing parameter selection,  window width,  62G05,  62E20,  62H99 
@article{1176350259,
     author = {Hall, Peter and Marron, J. S.},
     title = {On the Amount of Noise Inherent in Bandwidth Selection for a Kernel Density Estimator},
     journal = {Ann. Statist.},
     volume = {15},
     number = {1},
     year = {1987},
     pages = { 163-181},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176350259}
}

Hall, Peter; Marron, J. S. On the Amount of Noise Inherent in Bandwidth Selection for a Kernel Density Estimator. Ann. Statist., Tome 15 (1987) no. 1, pp.  163-181. http://gdmltest.u-ga.fr/item/1176350259/