Titlebook: Mathematical Statistics and Probability Theory; Volume B Statistical P. Bauer,F. Konecny,W. Wertz Conference proceedings 1987 D. Reidel Pub

兇兆

On Exponential Autoregressive Time Series Models,. Some of their models can be obtained from AREX models as special cases. The distribution of the innovation sequence (a probability mixture) and autoregressive structure of AREX processes are discussed as well.

規(guī)章

Estimating Quadratic Polynomials with Applications to Square Root Normalizing Transformations,e We study the behavior of a natural estimator of this and more general quadratic functions of normal parameters and obtain a necessary and sufficient condition for its admissibility under quadratic loss. In the case of inadmissibility, a class of better estimators is exhibited.

breadth

針葉類的樹

貧困

definition

熱情的我

OGLE

A Class of Non-Parametrically Constructed Parameter Estimators for a Stationary Autoregressive Modes are obtained by minimizing the functional. where. are respectively non-parametric estimators of the prediction function. and of f, the stationary initial density of (X.,…,X.). Consistency and asymptotic normality properties are proved.

Grating

The Extreme Linear Predictions of the Matrix-Valued Stationary Stochastic Processes,or of the linear predictions of matrix-valued stationary stochastic processes. It can be raised the question what happens if we apply another means of these diagonal elements. In this paper we use the geometric and harmonic means besides the arithmetic one for that purpose.

acrophobia

,Multivariate B — Splines, Analysis of Contingency Tables and Serial Correlation,om variables with all parameters integer except one, real are derived. It is shown how these formulae relate to B-splines and could be applied to compute the distribution of serial correlation coefficients and certain test statistics arising in the Bayesian analysis of contingency tables.