АНЫҚ ЕМЕС ЛОГИКА ЖӘНЕ КЛАСТЕРЛІК ТАЛДАУ ӘДІСТЕРІН ҚОЛДАНА ОТЫРЫП, КӘСІПОРЫНДАРДЫҢ ҚАРЖЫЛЫҚ ЖАҒДАЙЫН БОЛЖАУ

Abstract

Analysis of indicators of complexly structured economic systems operating under conditions of significant uncertainty, when there is no comprehensive statistics, or when non-financial data must be included in the indicators being studied, requires the use of special methodological tools for analyzing financial condition. One of these tools that has proven its adequacy for the problems being solved is the mathematical apparatus of the theory of fuzzy logic. The basis of the analytical support of any economic system is the methodology of financial analytics, which usually includes tools for recognizing and forecasting financial conditions under conditions of varying degrees of uncertainty. The article presents a methodology for analyzing the financial condition of an enterprise, which includes two sequentially performed procedures: 1) the formation of membership functions and 2) clustering in the space of indicators of financial conditions. As experimental studies have shown, the reliability of assignment to one of two clusters (normal and pre-bankruptcy state) is comparable to neural network methods. At the same time, the advantages of the technique remain due to the involvement of experts in compiling membership functions - this is the ability to provide a satisfactory recognition result in conditions of contradictory, incomplete data and various types of uncertainty factors.