Per-Johan Horgby, Ralf Lohse, and Nicola-Alexander Sittaro
Abstract
One of the most difficult issues in the medical underwriting of life insurance applicants is diabetes mellitus. Compiling the prognosticating parameters for diabetic applicants results in a complex system of mutually interacting factors. In addition, neither the prognosticating factors themselves nor their impact on the mortality risk is clear cut.
We show how a fuzzy inference system can be used in underwriting diabetes mellitus. A fuzzy inference system can cope with the imprecise nature of medical parameters by converting them into fuzzy sets and aggregating them using mathematical techniques. The fuzzy underwriting system presented goes further than previous applications of fuzzy set theory in insurance, as it is a real life application with contributions from insurance economics, insurance medicine, and computer science.
Key words and phrases: multiple risk factors, fuzzy inference, life insurance
Per-Johan Horgby,