Goals, as typically conceived in AI planning, provide an insufficient basis for choice of action, and hence are deficient as the sole expression of an agent's objectives. Decision-theoretic utilities offer a more adequate basis, yet lack many of the computational advantages of goals. We provide a preferential semantics for goals that grounds them in decision theory and preserves the validity of some, but not all, common goal operations performed in planning. This semantic account provides a criterion for verifying the design of goal-based planning strategies, thus providing a new framework for knowledge-level analysis of planning systems.
Although ceteris paribus preference statements concisely represent one natural class of preferences over outcomes or goals, many applications of such preferences require numeric utility function representations to achieve computational efficiency. We provide algorithms, complete for finite universes of binary features, for converting a set of qualitative ceteris paribus preferences into quantitative utility functions.
This article provides an overview of the field of qualitative decision theory: its motivating tasks and issues, its antecedents, and its prospects. Qualitative decision theory studies qualitative approaches to problems of decision making and their sound and effective reconciliation and integration with quantitative approaches. Although it inherits from a long tradition, the field offers a new focus on a number of important unanswered questions of common concern to AI, economics, law, psychology, and management. As developed by philosophers, economists, and mathematicians over some 300 years, these disciplines have developed many powerful ideas and techniques, which exert major influences over virtually all the biological, cognitive, and social sciences. Their uses range from providing mathematical foundations for microeconomics to daily application in a range of fields of practice, including finance, public policy, medicine, and now even automated device diagnosis.
This article provides an overview of the field of qualitative decision theory: its motivating tasks and issues, its antecedents, and its prospects. Qualitative decision theory studies qualitative approaches to problems of decision making and their sound and effective reconciliation and integration with quantitative approaches. Although it inherits from a long tradition, the field offers a new focus on a number of important unanswered questions of common concern to AI, economics, law, psychology, and management.
This paper provides an overview of recent work on qualitative approaches to decision theory. There is a good deal of basic work to be done on the ideas that are emerging from this work before we can apply them with much confidence in interactive problem solving. This abstract does not claim to do that work; it is meant to provide some references to the literature and a startingpoint for discussion. Disclaimer The 1997 AAAI Spring Symposium on qualitative decision theory revealed an emerging field with some promising ideas and potentially important applications. Unfortunately, there is still a large gap between the two. The most important trends that I can identify are still grappling in one way or another with the challenge of reworking a field that provides powerful theoretical arguments for representations of preferences that are not at all commonsensical, and that can be difficult to elicit. 1 At present, it is still unclear how to emerge from this foundational stage with an apparatus that will integrate with problem solving applications. The matter is complicated by the fact that there is no single dominant approach; different people have quite different ideas about how to proceed, and it may take a while for a few dominant paradigms to emerge. Even if the foundations were much clearer, I think it would probably be far from obvious how to integrate decision-theoretic and planning approaches to decision making. Decision theory (even qualitative decision theory) can produce preferences over courses of action, but these preferences (which in general will take into account and resolve a large number of competing evaluative factors, as well as considerations having to do 1I have in mind Savage's representation theorem in (Savage 1972), according to which the space of rational preferences over uncertain outcomes is isomorphic to a quantitative representation in terms of expected utility.