Модели, методы и информационные технологии разработки нечеткой экспертной системы диагностики финансового состояния предприятия

Abstract

Thesis for scientific degree of candidate of technical sciences, specialty 05.13.06 – information technologies. – National Technical University "Kharkiv Polytechnic Institute", Kharkov, 2016. The most important scientific and practical problem of identification of the financial condition of the company in terms of fuzzy data values controlled by financial indicators is considered in the thesis. Based on the analysis of traditional methods of solving this problem are identified the lack of them: insufficient consideration of the uncertainty of input data. In this regard, improved methods of multivariate discriminant analysis, cluster analysis and regression analysis for the case when the initial data – fuzzy numbers. For small sample output data suggested the procedure for building diagnostic information processing expert system that combines regression approach and Bayesian inference mechanism. The calculation of the regression coefficients method used pairwise comparisons. The method of predicting the time series correlated to the small sample of the original data, as well as a method of forecasting the series, given its wavelet model. On the basis of the developed methods based information technology diagnosis financial condition of the company. The proposed technology is integrated into the software package for the credit assessment of the bank borrowers showed more early detection of deterioration in the financial condition of borrowers compared to the old technology. The results obtained allowed to take timely management decisions and prevent increase of bad debts in the loan portfolio of the bank.