Propuesta de técnicas de minería de datos para la selección de atributos en la predicción del fracaso empresarial Proposal of data mining techniques for the selection of attributes in the prediction of business failure

Abstract

The analysis of financial information through the use of reasons has a limited capacity to quantify efficiently the financial success or failure of a company, due to the large number of indicators that can be used. The need arises to find indicators that make it possible to reveal, as soon as possible, insolvency or business failure processes and thus be able to evaluate possible scenarios for subsequent decision-making. In the prediction of business failure it is useful to have techniques that allow the selection of attributes, in this case financial indicators with high predictive power. The techniques of data mining are an appropriate proposal in the selection of these attributes with high predictive capacity in business failure processes. In this paper, the main limitations of financial analysis based on financial ratios are presented. The proposal of data miningtechniques that can be used in financial analysis for the prediction of business failure is based. In addition, the main advantages of data mining techniques in the decision- making process regarding the evaluation of business failure are discussed.