Abstract : Structural equation models (SEMs) are multivariate latent variable
models used to model causality structures in data. A Bayesian estimation and validation
of SEMs is proposed and identi ability of parameters is studied. The latter
study shows that latent variables should be standardized in the nalysis to ensure identi fiability. This heuristics is in fact introduced to deal with complex identi ability
constraints. To illustrate the point, identi ability constraints are calculated in a marketing application, in which posterior draws of the constraints are derived
from the posterior conditional distributions of parameters.
https://hal.archives-ouvertes.fr/hal-01125763
Contributor : Laboratoire Cedric <>
Submitted on : Sunday, March 22, 2020 - 7:34:35 PM Last modification on : Thursday, March 26, 2020 - 1:18:53 AM
Séverine Demeyer, Nicolas Fischer, Gilbert Saporta. Contributions to Bayesian Structural Equation Modeling. COMPSTAT'2010, 19th International Conference on Computational Statistics, Paris, Aug 2010, Paris, France. pp.469-476. ⟨hal-01125763⟩