Application of Bayesian networks to quantified risk assessment

This research investigates the possibility and benefits of applying Bayesian networks to quantified risk assessment. Bayesian networks are becoming an important formalism for representation of and inference with uncertain knowledge in artificial intelligence. We show that the central components of cause-consequence diagrams commonly used in quantified risk assessment can be expressed as Bayesian networks. Once such representation is established, the risk assessment using cause-consequence diagrams can be performed automatically using algorithms well developed in artificial intelligence and existing expert system shells. This will facilitate risk assessment of complex application domains. The method is illustrated using a well known example in risk assessment.