Results 1  10
of
17
Binary models for marginal independence
 JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B
, 2005
"... A number of authors have considered multivariate Gaussian models for marginal independence. In this paper we develop models for binary data with the same independence structure. The models can be parameterized based on Möbius inversion and maximum likelihood estimation can be performed using a versi ..."
Abstract

Cited by 16 (2 self)
 Add to MetaCart
A number of authors have considered multivariate Gaussian models for marginal independence. In this paper we develop models for binary data with the same independence structure. The models can be parameterized based on Möbius inversion and maximum likelihood estimation can be performed using a version of the Iterated Conditional Fitting algorithm. The approach is illustrated on a simple example. Relations to multivariate logistic and dependence ratio models are discussed.
A new algorithm for maximum likelihood estimation in Gaussian graphical models for marginal independence
 In U. Kjærulff and C. Meek (Eds.), Proceedings of the 19th Conference on Uncertainty in Artificial Intelligence
, 2003
"... Graphical models with bidirected edges (↔) represent marginal independence: the absence of an edge between two vertices indicates that the corresponding variables are marginally independent. In this paper, we consider maximum likelihood estimation in the case of continuous variables with a Gaussian ..."
Abstract

Cited by 15 (7 self)
 Add to MetaCart
Graphical models with bidirected edges (↔) represent marginal independence: the absence of an edge between two vertices indicates that the corresponding variables are marginally independent. In this paper, we consider maximum likelihood estimation in the case of continuous variables with a Gaussian joint distribution, sometimes termed a covariance graph model. We present a new fitting algorithm which exploits standard regression techniques and establish its convergence properties. Moreover, we contrast our procedure to existing estimation algorithms. 1
Mixed Effects Model Analyses of Incomplete Longitudinal . . .
, 1984
"... Incomplete longitudinal data are a common problem in clinical and epidemiological studies. This work was motivated by longitudinal studies of pulmonary function in young children characterized by both missing and mistimed observations as well as timevarying covariates. The objectives of this work w ..."
Abstract

Cited by 5 (2 self)
 Add to MetaCart
Incomplete longitudinal data are a common problem in clinical and epidemiological studies. This work was motivated by longitudinal studies of pulmonary function in young children characterized by both missing and mistimed observations as well as timevarying covariates. The objectives of this work were to develop a model and an analysis approach which 1) accommodated both missing and mistimed data and covariates which changed over time, 2) allowed for testing of hypotheses about both the fixed and random effects, and 3) were computationally feasible and practical. A generalized Mixed Effects Model was developed which generalized some assumptions used by previous authors. In particular, the restriction that the withinsubject variance is uncorrelated and homoscedastic (0'2 I) was generalized to the form 0'2Vi where Vi is any known positive definite matrix. Maximum Likelihood Estimators were derived and the EM algorithm and the Method of Scoring were used to solve the maximum likelihood equations. Randomly generated data were used in a preliminary exploration of the
The Wishart Distributions on Homogeneous Cones
, 2001
"... The classical family of Wishart distributions on a cone of positive definite matrices and its fundamental features are extended to a family of generalized Wishart distributions on a homogeneous cone using the theory of exponential families. The generalized Wishart distributions include all known fam ..."
Abstract

Cited by 4 (0 self)
 Add to MetaCart
The classical family of Wishart distributions on a cone of positive definite matrices and its fundamental features are extended to a family of generalized Wishart distributions on a homogeneous cone using the theory of exponential families. The generalized Wishart distributions include all known families of Wishart distributions as special cases. The relations to graphical models and Bayesian statistics are indicated.
An Evaluation of Some Approximate F Statistics and Their Small Sample Distributions for the Mixed Model with . . .
, 1987
"... The purpose of this work was to extend results from the General Linear Univariate Model and the General Linear Multivariate Model to special cases of the mixed model with linear covariance structure. These extensions were then used to motivate approximate F statistics for the mixed model. Three appr ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
The purpose of this work was to extend results from the General Linear Univariate Model and the General Linear Multivariate Model to special cases of the mixed model with linear covariance structure. These extensions were then used to motivate approximate F statistics for the mixed model. Three approximate F statistics were proposed; one was based on the canonical form of the mixed model (FREML) and two were based on weighted least squares (F WLS ' F
Optimal Decomposition and Classification of Linear Mixtures of ARMA Processes
 In preparation
"... : We consider the problem of detecting and classifying an unknown number of multiple simultaneous Gaussian autoregressivemoving average (ARMA) signals with unknown variances given a finite length observation of their sum and a dictionary of candidate ARMA models. We introduce the concept of mixture ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
: We consider the problem of detecting and classifying an unknown number of multiple simultaneous Gaussian autoregressivemoving average (ARMA) signals with unknown variances given a finite length observation of their sum and a dictionary of candidate ARMA models. We introduce the concept of mixture of ARMA (\SigmaARMA) process and show that the detection and classification problem is equivalent to decomposition of a \SigmaARMA process. The optimal solution in an information theoretic sense is found by applying the minimum description length (MDL) principle. Computation of the MDL solution requires the maximization of a likelihood function of the variances of the ARMA components for a series of subsets from the dictionary. This maximization is performed efficiently by combining the EM algorithm with the RauchTungStriebel optimal smoother or with a Wiener filterbased approximation thereof and a new result on subset selection. The performance of the algorithm is illustrated by numeric...
Restricted concentration models  graphical Gaussian models with
"... In this paper we introduce restricted concentration models (RCMs) as a class of graphical models for the multivariate Gaussian distribution in which some elements of the concentration matrix are restricted to being identical is introduced. An estimation algorithm for RCMs, which is guaranteed ..."
Abstract
 Add to MetaCart
In this paper we introduce restricted concentration models (RCMs) as a class of graphical models for the multivariate Gaussian distribution in which some elements of the concentration matrix are restricted to being identical is introduced. An estimation algorithm for RCMs, which is guaranteed to converge to the maximum likelihood estimate, is presented.
The Analysis of CrossSectional Time
, 1975
"... This study is concerned with the estimation of linear relationships from crosssectional time series data. The subject has been extensively discussed in the econometric literature. The diversity of the approaches proposed in the literature stems both from the different sets of assumptions and the di ..."
Abstract
 Add to MetaCart
This study is concerned with the estimation of linear relationships from crosssectional time series data. The subject has been extensively discussed in the econometric literature. The diversity of the approaches proposed in the literature stems both from the different sets of assumptions and the different estimation procedures adopted. Most of the alternative approaches are based on simpler assumptions than the more realistic assumptions used here. The variance component approaches ignore the possibility of serial· correlation in the time,direction. The seemingly unrelated regressions approaches aSSume q