Knowledge discovery in medical databases based on rough sets and attribute-oriented generalization

This paper presents a knowledge discovery system based on rough sets and attribute-oriented generalization and its application to medicine. Diagnostic rules and information on attributes are extracted from clinical databases on diseases of congenital anomaly. The induced results show that this method extracts experts' knowledge correctly and it also discovers that symptoms observed in six positions (eyes, noses, ears, lips, fingers and feet) play important roles in differential diagnosis.