DATA ANALYSIS OF THE FINANCIAL INDICATORS OF POLISH COMPANIES

Abstract

This article aims to present the issue of the company ́s bankruptcy and defines which financial indicators affects and can accurately detect the financial health of the company and thus better predict the emergence of potential bankruptcy. Currently, these methods include mainly modern techniques from the data mining area. For the practical application of this approach to predict the future state of the company, were used the financial indicators of Polish companies. We used the most suitable algorithms for predicting bankruptcies – decision trees that provide simple results interpretation. The analytical process is managed by the CRIPS-DM methodology, which offers a description of the important steps needed to solve this task. Part of the article constitutes an analysis of the current state, which presents solutions to this problem by other authors. Analysing available data we found that the most effective financial indicators are Attr27[profit on operating activities / financial expenses], Attr34 [operating expense/total liabilities] and Attr41 (total liabilities/[(profit on operating activities + depreciation)*(12*365)]). The model that best-predicted bankruptcy was the C5.0 decision tree algorithm.