Hsinyen Wei and Costas Tsatsoulis
We propose a new methodological approach to CBR that allows it to use decision theoretic approaches to deal with multiple types of uncertainty. The retrieval of old cases in CBR is viewed as a decision problem, where each case from the case base provides an alternative solution and a prediction for the possible outcomes for the current problem. When uncertainty is encountered during case-based problem solving, decision theory is applied to evaluate each potential case in terms of the attributes that are significant for the current problem, so that the most desirable old case can be selected. Such integration provides a perfect complement between CBI~ and decision analysis.