Application of Rough Sets in diagnosis of the depressive state of mind

Rough Set Theory is an emerging rule based soft computing methodology that employs approximations of crisp concepts. It has been used widely for knowledge discovery in real life data-centric applications that typically include uncertain or incomplete data. This paper describes an application of rough sets in identifying depressive episodes in the field of psychiatry. The core concepts of rough sets such as Reduct and Core are used to reduce the number of descriptive attributes based on their relative significance. The reduced information system yields a compact set of high-strength rules that identify the state of mind of a person to categorize new patients with high accuracy. We illustrate how Rough Sets can find symbolic and easily readable rules that could be used fruitfully by psychiatrists for clinical diagnosis.