A Learning Apprentice for Decision Support Systems

In this paper we present a general Learning Apprentice linked to decision support systems for generating diagnoses. The Learning Apprentice acquires diagnostic knowledge by following the case studies and exemplars presented to the decision support system. It employs three machine learning techniques to learn: Exemplar Based Learning is used to acquire new diagnosis candidates, Explanation Based Learning and Similarity Based Learning are used to specialize known diagnoses or discriminate against other diagnoses. We also outline the architecture of a generic decision support system which generates a diagnosis based on the current data entered by the user.