Decision making by means of probabilistic sets in medical expert systems

Decision-making problems in expert systems often involve an aggregation of information gathered in different nodes of decision trees (graphs, networks). This information can originate from anamnesis, physical examination, or medical tests of the patient under control of the physician and expert system. Such information (divided into various groups) usually has to be aggregated in order to draw conclusions indicating which branch (arc) of the decision tree (graph) should be chosen for moving from one state to the other. An approach to the solution of decision problems in expert systems is presented that derives from multicriteria decision making by means of probabilistic set theory.<>