Aplicação da Teoria dos Rough Sets na Previsâo do Fracasso Empresarial

Abstract

[Abstract]: Prediction of corporate bankruptcy is a phenomenon of increasing interest to investors, creditors, borrowing firms, and governments alike. In the literature of the specialty we can found a lot of works that apply a wide variety of techniques to analyze this problem. In this work we present a method relatively recent of evaluation of the risk of business failure with base of Rough Sets theory. This theory allows obtaining a group of condition attributes that assure an acceptable level of classification of the analyzed companies, allowing with that to obtain rules to elaborate a decision making process system. For the selection of the subset of the relevant attributes was used the minimum description length principle. For the empiric application we have used a sample of Portuguese companies of mixed industries. The overall predictive accuracy of the model for one, two, and three years before bankruptcy was 97,43%, 93,15% and 91,69%, respectively.