NOTE ON ASYMPTOTIC NORMALITY OF KERNEL DENSITYESTIMATOR FOR LINEAR PROCESS UNDER SHORT-RANGEDEPENDENCE
Konrad Furmańczyk
Abstract: We consider the problem of density estimation for a one-sided linear process
with i.i.d. square integrable innovations We prove that
under weak conditions on which imply short-range dependence of the linear
process, finite-dimensional distributions of kernel density estimate are asymptotically
multivariate normal. In particular, the result holds for with which is
much weaker than previously known sufficient conditions for asymptotic normality. No
conditions on bandwidths are assumed besides and The proof uses
Chanda’s [1], [2] conditioning technique as well as Bernstein’s “large block-small block”
argument.