A case-based model for sorting problems in multiple criteria decision analysis

Sorting is a process for multiple criteria decision analysis that arranges a set of alternatives into ordinally defined groups. Based on weighted Euclidean distance, a case-based distance model is proposed for sorting using weights and group thresholds obtained by assessment of a case set. A quadratic optimization program is employed to find optimal weights and group thresholds. Instead of direct inquiry, this case-based method can elicit the decision maker´s preferences more expeditiously and accurately. A case study on Canadian municipal water usage analysis is presented.