ERUDIT-Workshop
Fuzzy Diagnostic and Therapeutic Decision Support
11–12 May 2000, Vienna, Austria
Department of Medical Computer Sciences, Section on Medical Expert and Knowledge-Based Systems´
University of Vienna Medical School, Spitalgasse 23, A-1090 Vienna, Austria
e-mail: michael.schuerz@akh-wien.ac.at
Abstract. Defining fuzzy concepts in medicine is one crucial point in designing medical expert systems. In this paper, three methods are presented for learning fuzzy membership functions semi-automatically from data. These methods apply descriptive statistics, artificial neural networks, and least square curve fitting. Moreover, an evaluation was carried out to compare medically defined laboratory reference ranges with the obtained results. Among those least square curve fitting seems to be the method with the most suitable results.
(In: Adlassnig, K.-P. (ed.) Fuzzy Diagnostic and Therapeutic Decision Support. Proc. of the ERUDIT Workshop, Österreichische Computer Gesellschaft, Vienna, p.129.)