Issue No. 04 - July-August (1997 vol. 12)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/64.608193
<p>In most realistic problem-solving activities, the problem solver faces two major issues: how to deal with unknown problem features, and how to make decisions in the presence of these unknowns. We've developed a methodology that lets case-based reasoning use decision-theoretic approaches to deal with these two issues. We view CBR as a technology for automated, intelligent problem solving; the goal of integrating CBR and decision theory is to improve the ability of CBR systems to solve problems in domains of incomplete information. (See the sidebars for more information on CBR and decision theory.) </p> <p>Our methodology views the retrieval of old cases in CBR as a decision problem, where each case from the case base provides an alternative solution and a prediction of the possible outcomes for the problem. When case-based problem solving encounters uncertainty, our methodology applies decision theory to evaluate each case in terms of the attributes that are significant for the problem, so that the most desirable case can be selected. </p> <p>We implemented our methodology in a case-based design assistant that helps chemists design pharmaceuticals. The system proposes chemicals for generating drugs and can evaluate the various design choices. Drug design is an appropriate application domain. The number of possible compounds that must be explored during the design phase is enormous. Also, the evaluation of design choices is extremely important, because it lets the chemist focus on a small subset of such compounds. Drug development is a very difficult design task where an intelligent assistant can greatly enhance the quality of the compounds generated and improve the chemist's productivity. The interactions of chemicals or their effect on the human body are often not known, and the specifications are necessarily incomplete. Finally, a compound might have multiple effects, both positive and negative. Its usefulness in a specific situation needs to be evaluated by weighing the drug's utility to the target population versus the predicted risks. </p>
Q. Cheng, C. Tsatsoulis and H. Wei, "Integrating Case-Based Reasoning and Decision Theory," in IEEE Intelligent Systems, vol. 12, no. , pp. 46-55, 1997.