Various special linear structures connected with covariance matrices are reviewed and graphical methods for their representation introduced, involving in particular two different kinds of edge between the nodes representing the component variables. The distinction between decomposable and nondecomposable structures is emphasized. Empirical examples are described for the main possibilities with four component variables.
Publié le : 1993-08-14
Classification:
Chain model,
conditional independence,
covariance selection,
decomposable model,
linear structural equation,
multivariate analysis,
path analysis
@article{1177010887,
author = {Cox, D. R. and Wermuth, Nanny},
title = {Linear Dependencies Represented by Chain Graphs},
journal = {Statist. Sci.},
volume = {8},
number = {4},
year = {1993},
pages = { 204-218},
language = {en},
url = {http://dml.mathdoc.fr/item/1177010887}
}
Cox, D. R.; Wermuth, Nanny. Linear Dependencies Represented by Chain Graphs. Statist. Sci., Tome 8 (1993) no. 4, pp. 204-218. http://gdmltest.u-ga.fr/item/1177010887/