×

The selection of window widths in kernel nonparametric regression. (English) Zbl 0602.62032

The estimation of the regression function \(r(x)=E[Y| X=x]\) using kernel density estimation techniques has long been hindered by reliance on subjective techniques for determining the window width required. Recent work in data-based optimal selection of nonparametric density estimates has eliminated the subjectivity of window selection for density estimation, but less attention has been given to the specialized nature of the kernel regression estimators.
In this work, we demonstrate intuitively and empirically that a certain optimal bivariate window width is to be preferred over optimal univariate window widths in forming kernel estimators of r(x). Simulations also reveal how well data-based optimal kernel regression estimates perform for simple linear, polynomial, and exponential models.

MSC:

62G05 Nonparametric estimation
65C99 Probabilistic methods, stochastic differential equations
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Cheng K.F., Annals of the Institute of Statistical Mathematics 32 pp 187– (1980) · Zbl 0459.62039 · doi:10.1007/BF02480324
[2] Epanechnikov V., Theory of Probability and Its Applications 14 pp 153– (1969) · Zbl 0175.17101 · doi:10.1137/1114019
[3] Hall P., Annals of Statistics 11 pp 1156– (1983)
[4] International Mathematical and Statistical Library,Reference Manual (1980)
[5] Rice J., Annals of Statistics 12 pp 1215– (1984) · Zbl 0554.62035 · doi:10.1214/aos/1176346788
[6] Rudemo M., Scandinavian Journal of Statistics 9 pp 65– (1982) · Zbl 0501.62028
[7] Wahba G., Communications in Statistics 4 pp 1– (1975)
[8] Watson G.S., Sankhya 26 pp 358– (1964)
[9] Woodfield T.J. Three data-based approaches to bivariate nonparametric density estimation based upon MISE criteria 1985 Submitted
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.