Qualitative models and Application to Medical Diagnosis

by Liliana Ironi



in Proc. 9th France-URSS-Italy Joint Symposium in Computational Mathematics and Applications, INRIA-Sophia Antipolis, (1991), 139-157.


ABSTRACT

Current research in Artificial Intelligence is attempting to develop a framework for the exploitation of model-based reasoning in a diagnostic environment. In some application domains, for example the medical one, where the knowledge of values of parameters and functional relationships about variables is often incomplete, a qualitative formulation seems to be more adequate than a quantitative one because of the reduced number of assumptions in building the model. Moreover, for many problems, there might be interest only in few features, often of qualitative nature, of the solution. After a brief overview of the existing approaches to qualitative modeling, the {\bib Qsim} method is described. Finally, it is discussed how qualitative models and their simulation results can be exploited in a medical diagnostic expert system in order to highly enhance its performance.




  • Back to publication list


    Liliana Ironi June 25th, 1996