A possibilistic formalization of case-based reasoning and decision making

The idea of case-based decision making has recently emerged as a new paradigm for decision making under uncertainty. It combines principles from decision theory and case-based reasoning, a problem solving method in artificial intelligence. In this paper, we propose a formalization of case-based reasoning which is based on possibility theory and utilizes approximate reasoning techniques. The corresponding approach to case-based decision making is realized as a two-stage process. In the first stage, the decision maker applies case-based reasoning in order to quantify the uncertainty associated with different decisions in form of possibility distributions on the set of consequences. In the second stage, generalizations of expected utility theory are used for choosing among acts resp. the associated distributions.