Beyond Z-Analysis: Self Organizing Neural Networks for Financial Diagnosis

Abstract

In this paper we propose a complete method for financial diagnosis based on Self Organizing Feature Maps (SOFM), a neural network model which, on the basis of the information contained in a multidimensional space--in our case, financial ratios--generates a space of lesser dimensions. In this way, similar input patterns are represented close to one another on a map. The methodology has been complemented and compared with multivariate statistical models such as Linear Discriminant Analysis (LDA), as well as with neutral models such as the Multilayer Perception (MLP). As the principal advantage which distinguishes the proposed methodology from other statistical techniques that have been developed to analyze accounting information, mention should be made of its robustness in not demanding that the input variables follow any distribution function, thus providing a complete analysis which goes beyond that of the traditional models based on Z score, without renouncing simplicity for the final decision maker.

JEL Classification: G00