This paper develops the theory of density estimation on the Stiefel ma
nifold V-k,V-m where V-k,V-m is represented by the set of m x k matric
es X such that X'X= I-k, the kxk identity matrix. The density estimati
on by the method of kernels is considered. proposing two classes of ke
rnel density estimators with small smoothing parameter matrices and fo
r kernel functions of matrix argument. Asymptotic behavior of various
statistical measures of the kernel density estimators is investigated
ibr small smoothing parameter matrix and/or for large sample size. Som
e decompositions of the Stiefel manifold V-k,V-m play useful roles in
the investigation, and the general discussion is applied and examined
for a special kernel function. Alternative methods of density estimati
on are suggested, using decompositions. (C) 1998 Academic Press.