Empirical Process of Multivariate Gaussian under General Dependency
share
This paper explores certain kinds of empirical process with respect to the components of multivariate Gaussian. We put forward some finite sample bounds which hold for multivariate Gaussian under general dependency. As a direct corollary, we prove that the empirical distribution of a Gaussian process will converge, that is to say, sup_t |F_n(t) - EF_n(t) |P 0, as long as the covariance of the Gaussian process vanishes with the time shift.
READ FULL TEXT