The rate of convergence of various sweep strategies of stochastic relaxation for simulating multivariate Gaussian measures are calculated and compared. Each sweep strategy prescribes a method for chosing which coordinates of the random vector are to be updated. Deterministic sweep strategies in...

In the class of multivariate exponential power distributions, we derive LBI (locally best invariant) tests for normality in the two cases: (i) mean vector [mu] is known and (ii) [mu] is unknown. In the case (i), the null and nonnull asymptotic distributions of the test statistic are derived. In...

Suppose on a probability space ([Omega], F, P), a partially observable random process (xt, yt), t = 0; is given where only the second component (yt) is observed. Furthermore assume that (xt, yt) satisfy the following system of stochastic differential equations driven by independent Wiener...

Some equivalence conditions for an asymptotically mean stationary channel to be ergodic are given. Incidental results on asymptotically mean stationary measures are also given. Absolute continuity of measures plays a crucial role.

We give an example of a central Wishart matrix W with one degree of freedom and scale matrix of rank 2 such that the diagonal entries of W are not associated. This allows us to conclude that no central Wishart matrix with one degree of freedom and scale matrix of rank greater than 1 is...

Capacity is deremined for a class of communication channels containing additive noise. Gaussian noise processes and a large class of non-Gaussian processes are included. The constraint on the transmitted signals is given in terms of an increasing family of finite-dimensional subspaces. For this...

An asymptotic distribution theory is developed for a general class of signed-rank serial statistics, and is then used to derive asymptotically locally optimal tests (in the maximin sense) for testing an ARMA model against other ARMA models. Special cases yield Fisher-Yates, van der Waerden, and...

For the mean vector of a p-variate normal distribution (p [greater, double equals] 3), the generalized Bayes estimators dominating the James-Stein estimator under quadratic loss are given based on the methods of Brown, Brewster and Zidek for estimating a normal variance.

We propose a natural test of fit of a parametric regression model. The test is based on a comparison of a nonparametric kernel estimate of a regression function with its least-squares parametric estimate. Under the null hypothesis we derive approximations to the probability distribution...

The main purpose of this paper is to investigate high dimensional limiting behaviors, as m becomes infinite (m -- [infinity]), of matrix statistics on the Stiefel manifold Vk, m, which consists of m - k (m = k) matrices X such that X'X = Ik. The results extend those of Watson. Let X be a random...