Different aspects of Gaussian Bayesian networks

R. Susi García, M. A. Gómez Villegas

A Bayesian network (BN) is a probabilistic graphical model used to study a set of variables with a known probabilistic dependence structure. This graphical model has become a necessary tool for reasoning under uncertainty in complex domains. Nevertheless, building a BN is a complicate task because it requires the speciation of a large number of parameters subject to biases. Therefore, sensitivity analyses are necessary to evaluate the effect of uncertainty in the network and to determine the values of the parameters to get accurate network outputs.
In this work, we focus on a subclass of BN known as Gaussian Bayesian network (GBN). GBNs encode probabilistic relationships among continuous variables with Gaussian distributions. Our aim is to introduce different aspects of sensitivity analysis in GBN that are still open.

Palabras clave: Gaussian Bayesian network, sensitivity analysis, sensitivity measure

L08.2 Grupo de Inferencia Bayesiana: Applications of Bayesian Modelling in Econometrics and Genetics
5 de septiembre de 2016  15:40
0.09 - Aula de proyectos 2

M. E. Castellanos Nueda, C. Armero, S. Cabras, A. Quirós Carretero

L. Sanz San Miguel, M. A. Gómez Villegas, J. M. Marín Diazaraque

M. Wiper, P. Galeano, C. García de la Fuente