-
作者：KELLY, RJ; MATHEW, T
作者单位：University System of Maryland; University of Maryland Baltimore County
摘要：A linear model with two variance components is considered, one variance component (say, sigma1(2) greater-than-or-equal-to 0) corresponding to a random effect, and a second variance component (say, sigma2 > 0) corresponding to the experimental errors. A class of invariant quadratic estimators (IQEs) is characterized, having uniformly smaller mean-squared error (MSE), and uniformly smaller probability of negativity, compared with the analysis-of-variance (ANOVA) estimator of sigma1(2). It turns...
-
作者：VOVK, VG
摘要：A simple example shows that the classical theory of probability implies more than one can deduce via Kolmogorov's calculus of probability. Developing Dawid's ideas I propose a new calculus of probability which is free from this drawback. This calculus naturally leads to a new interpretation of probability. I argue that attempts to create a general empirical theory of probability should be abandoned and we should content ourselves with the logic of probability establishing relations between pro...
-
作者：HILL, BM
作者单位：University of Michigan System; University of Michigan
摘要：A class of parametric models, called splitting processes, is defined, by using de Finetti's concept of adherent mass. Such splitting processes give rise to complex mixtures of distributions. It is proved that the nonparametric Bayesian predictive procedure A(n), of Hill, holds exactly for a member of this class called a nested splitting process. The connection between A(n) and the Dirichlet process is stated and proved. A multivariate version of A(n), based on splitting processes, is proposed.
-
作者：WELLS, MT; JAMMALAMADAKA, SR; TIWARI, RC
作者单位：University of California System; University of California Santa Barbara; University of North Carolina; University of North Carolina Chapel Hill
摘要：Let X1, . . ., X(n) be a sequence of independent and identically distributed random variables with an unknown underlying continuous cumulative distribution function F. Often we would like to test a null hypothesis concerning the goodness of fit of F to some distribution function which is fully specified or belongs to some parametric family, i.e. in some applications the null hypothesis is simple whereas in others it may be composite. In this paper we present the large sample theory of tests ba...
-
作者：QUEEN, CM; SMITH, JQ
作者单位：University of Warwick
摘要：Multiregression dynamic models are defined to preserve certain conditional independence structures over time across a multivariate time series. They are non-Gaussian and yet they can often be updated in closed form. The first two moments of their one-step-ahead forecast distribution can be easily calculated. Furthermore, they can be built to contain all the features of the univariate dynamic linear model and promise more efficient identification of causal structures in a time series than has b...
-
作者：KUNERT, J; UTZIG, BP
作者单位：Free University of Berlin
摘要：There is concern that the usual analysis of crossover designs with more than two treatments is subject to bias due to correlations between the measurements on the same experimental units. It has been shown by Kunert in the special case of balanced Latin squares that this bias can be present but that it is limited. Extending the work of Kunert we show in the present paper that there is a constant X* which guarantees that the estimate for the variance of any treatment contrast from the usual mod...
-
作者：WEST, M
摘要：Kernel density estimation techniques are used to smooth simulated samples from importance sampling function approximations to posterior distributions, resulting in revised approximations that are mixtures of standard parametric forms, usually multivariate normal or T-distributions. Adaptive refinement of such mixture approximations involves repeating this process to home-in successively on the posterior. In fairly low dimensional problems, this provides a general and automatic method of approx...
-
作者：SUNDBERG, R
摘要：We demonstrate the close relationship between first-factor continuum regression and standard ridge regression. The difference is that continuum regression inserts a scalar compensation factor for that part of the shrinkage in ridge regression that has no connection with tendencies towards collinearity. We interpret this to mean that first-factor continuum regression is preferable in principle to ridge regression if we want protection against near collinearity but do not admit shrinkage as a ge...
-
作者：VARDI, Y; LEE, D
作者单位：Nokia Corporation; Nokia Bell Labs; AT&T
摘要：The problem of recovering an input signal from a blurred output, in an input-output system with linear distortion, is ubiquitous in science and technology. When the blurred output is not degraded by statistical noise the problem is entirely deterministic And amounts to a mathematical inversion of a linear system with positive parameters, Subject to positivity constraints on the solution. We show that all such linear inverse problems with positivity restrictions (LININPOS problems for short) ca...
-
作者：GILKS, WR; CLAYTON, DG; SPIEGELHALTER, DJ; BEST, NG; MCNEIL, AJ; SHARPLES, LD; KIRBY, AJ
摘要：We review applications of Gibbs sampling in medicine, involving longitudinal, spatial, covariate measurement and survival models. Applications in immunology, pharmacology, transplantation, cancer screening, industrial epidemiology and genetic epidemiology are discussed.
