Bankruptcy Prediction with Rough Sets

Abstract

textabstractThe bankruptcy prediction problem can be considered an or\ndinal classification problem. The classical theory of Rough Sets describes\nobjects by discrete attributes, and does not take into account the order-\ning of the attributes values. This paper proposes a modification of the\nRough Set approach applicable to monotone datasets. We introduce re-\nspectively the concepts of monotone discernibility matrix and monotone\n(object) reduct. Furthermore, we use the theory of monotone discrete\nfunctions developed earlier by the first author to represent and to com-\npute decision rules. In particular we use monotone extensions, decision\nlists and dualization to compute classification rules that cover the whole\ninput space. The theory is applied to the bankruptcy prediction problem.