Decision Making in Qualitative Influence Diagrams
Silja Renooij, Linda C. van der Gaag · 1998
The increasing number of knowledge-based systems that build on a Bayesian belief network or influence diagram acknowledge the usefulness of these frameworks for addressing complex real-life problems. The usually large number of probabilities and utilities required for their application, however, is often considered a major obstacle. The use of qualitative abstractions may to some extent remove this obstacle. Qualitative belief networks and associated algorithms have been developed before. In this paper, we address qualitative influence diagrams and outline an e#cient algorithm for qualitative decision making. 1 Introduction In the late 1980s, the framework of Bayesian belief networks was introduced for reasoning with uncertainty [Pearl 1988]. The framework provides a formalism for encoding a joint probability distribution on a set of statistical variables and o#ers algorithms for probabilistic inference. In practice, reasoning with uncertainty is often performed to support a decision ...