Time-Independent Rule-Based Guideline Induction

Whereas decision trees are widely studied in the context of knowledge representation and inductive learning, guidelines can be considered as an extension of decision trees that have not received the same amount of attention in the artificial intelligence community, even though they have been proved highly useful in several decision support problems as the therapy process in health-care. This paper is about the formal definition of guidelines as a rulebased knowledge structure and also about the introduction of an incremental inductive learning algorithm to develop and update timeindependent guidelines. These ideas have been tested in the healthcare domain of some heart diseases.