Probabilistic inference in influence diagrams

This paper studies the relationship between probabilistic inference in Bayesian networks and evaluation of influence diagrams. We clearly identify and separate from other computations probabilistic inference subproblems that must be solved in order to evaluate an influence diagram. This work leads to a new method for evaluating influence diagrams where arbitrary Bayesian network inference algorithms can be used for probabilistic inference. We argue that if the VE algorithm (Zhang and Poole 1994, Zhang and Poole 1996) is used for probabilistic inference, the new method is more efficient than the best previous methods (Shenoy 1992, Jensen et al 1994). If a more efficient probabilistic inference algorithm such as the VEC algorithm (Zhang and Poole 1996) or some approximation algorithm is used, the new method can be even more efficient