(a) For the bandwidth matrix selectors, there are several varieties: (i) plug-in
Hpi, (ii) least squares (or unbiased) cross validation
(LSCV or UCV)
Hlscv, (iii) biased cross validation (BCV)
Hbcv and (iv) smoothed cross validation (SCV)
Hscv. Scalar bandwidth selectors are not provided - see
sm or KernSmooth packages.
(b) For kernel density estimation, the main function is
kde. For kernel discriminant analysis, it's kda.kde.
(c) For display, versions of plot send to a graphics window
the results of density estimation or discriminant analysis.
For d = 1, 2, 3, 4, binned kernel estimation is available from the
KernSmooth library.
For an overview of this package with 2-dim density estimation, see
vignette("kde").
Sain, S.R., Baggerly, K.A. & Scott, D.W. (1994) Cross-validation of multivariate densities. Journal of the American Statistical Association. 82, 1131-1146.
Scott, D.W. (1992) Multivariate Density Estimation: Theory, Practice, and Visualization. John Wiley & Sons. New York. Simonoff, J. S. (1996) Smoothing Methods in Statistics. Springer-Verlag. New York.
Wand, M.P. & Jones, M.C. (1994) Multivariate plugin bandwidth selection. Computational Statistics 9, 97-116. Wand, M.P. & Jones, M.C. (1995) Kernel Smoothing. Chapman & Hall/CRC. London.
sm, KernSmooth