Graphical Models: Applicability and Software
نویسندگان
چکیده
A class of models for mixed continuous and nominal data was proposed by Lauritzen and Wermuth. Log-linear models, linear regression and (M)ANOVA are special cases of these mixed graphical models. Their clear interpretation by means of independence graphs based on Markov properties makes them useful in applied multivariate statistics. Only few statistical software is developed yet for these models. MIM and CoCo are brieey discussed in this paper. XLISP-STAT seems to be an attractive environment for further software development in this area.
منابع مشابه
APG: An Efficient Software Program for Amp-Pl Thermobarometry Based on Graphical Method
Geothermobarometry equations are based on thermodynamic principles and appear in single or multi-variant functions. The number of variants for a specific composition or reaction usually is reduced into 2 involving temperature (T) and pressure (P). Since most of planned equations have two passive or variant P and T, using these equations should be with special care. It is very effective to use g...
متن کاملCopula Gaussian Graphical Models *
We propose a comprehensive Bayesian approach for graphical model determination in observational studies that can accommodate binary, ordinal or continuous variables simultaneously. Our new models are called copula Gaussian graphical models and embed graphical model selection inside a semiparametric Gaussian copula. The domain of applicability of our methods is very broad and encompass many stud...
متن کاملBehaviour-Preserving Transformations in SHE: A Formal Approach to Architecture Design
SHE (Software/Hardware Engineering) is an objectoriented analysis, specification and design method for complex reactive hardware/software systems. SHE is based on the formal specification language POOSL and a design framework guiding the analysis and design activities. During analysis several graphical models are produced. These models are formalized in a textual representation in the POOSL lan...
متن کاملCopula Gaussian Graphical Models and Their Application to Modeling Functional Disability
We propose a comprehensive Bayesian approach for graphical model determination in observational studies that can accommodate binary, ordinal or continuous variables simultaneously. Our new models are called copula Gaussian graphical models (CGGMs) and embed graphical model selection inside a semiparametric Gaussian copula. The domain of applicability of our methods is very broad and encompasses...
متن کاملCopula Gaussian Graphical Models and Their Application to Modeling Functional Disability Data1 by Adrian Dobra
We propose a comprehensive Bayesian approach for graphical model determination in observational studies that can accommodate binary, ordinal or continuous variables simultaneously. Our new models are called copula Gaussian graphical models (CGGMs) and embed graphical model selection inside a semiparametric Gaussian copula. The domain of applicability of our methods is very broad and encompasses...
متن کامل