Systematic Creation of Bayesian Belief Networks: Relationship with Logic Logic-Evolved Decision Tools

摘要

Bayesian Belief Networks (BBNs) are rapidly becoming a popular modeling technique employed by decision support systems. While BBNs can be effectively implemented to represent countless probabilistically or causally related real world concerns, the systematic development of these graphical structures is significantly lacking. Potential benefits of adhering to an organized approach throughout the belief network construction process are forfeited because existing standardized procedures for this task are executed infrequently. This paper addresses a framework for an orderly development of a BBN and explores relationships between capabilities of the Logic-^sEvolved Decision (LED) Tools software package and the automatic construction of BBNs.