Bankruptcy prediction using a data envelopment analysis

There is a growing recognition that a variety of machine learning problems can be approached advantageously by tools from the field of optimization. In this paper we compare the classification performance of a linear programming model, a data envelopment (DEA) model and a rule induction (C5.0) model. In terms of accuracy and employment the DEA model outperforms the other models. (C) 2003 Elsevier B.V. All rights reserved.