Using Set Covering and Uncertain Reasoning to Determine Treatments

Abstract Reggia's generalized set covering model is for determining the diseases which are present in a patient from his symptoms. In the current paper, several shortcomings in the basic model are described, and it is shown how these shortcomings can be overcome by employing a method for reasoning under uncertainty. Secondly, the model is used to determine treatments for a stroke patient from the diseases present, showing that set covering is appropriate for determining treatments from diseases as well as diseases from symptoms.