Title: A Computer Algebra-Based Expert System for Diagnosis of Anorexia

Authors: Cristina Perez-Carretero (1)
         Luis M. Laita            (1)
         Eugenio Roanes-Lozano    (2)
         Luis Lazaro              (3)
         Jesus Gonzalez-Cajal     (3,4)

         (1) Dept. Artificial Intelligence, Univ. Politecnica Madrid
         (2) Dept. Algebra, Univ. Complutense Madrid
         (3) C.U. Cardenal Cisneros
         (4) Hospital de la Princesa

Abstract:

This paper deals with the construction of a Logic and Computer Algebra-based
expert system devoted to diagnosis of anorexia. Predisposition, triggering
and maintenance factors, together with symptoms of the illness, behave as
logical antecedents of production rules. These production rules lead,
through the partial measure of the intensity of the factors and symptoms,
to a final diagnosis. The inference engine works with polynomial expressions
which are the translation of the production rules. The inference processes
are translated to ideal membership problems.