Логистический нейросетевой метод построения динамических моделей банкротств при неполных данных

Горбатков Станислав Анатольевич, Белолипцев Илья Игоревич, Фархиева Светлана Анатольевна · Интернет-журнал Науковедение · 2014

Hitherto not been investigated Neural network dynamic models estimating the probability of the risk of bankruptcy are of great practical interest today. But they have not been sufficiently investigated yet. Knowledge of developmental dynamics of the bankruptcy process allows the manager to determine the stage of bankruptcy and to apply proactive control actions. However, the creation of dynamic models of the bankruptcy risk is complicated by the presence of incomplete data. In this paper, we proposed an original hybrid neural network iterative method for estimating the probability of developing in time the risk of bankruptcy of enterprises. The method is focused on the using in difficult conditions: the incompleteness of the data and their strong noising. Computational experiments have confirmed the effectiveness of the basic idea of creating a dynamic model of bankruptcies with continuous time. The developed method can be used in a wide range of applications: assessing the creditworthiness of the borrower, economic security by continually monitoring their financial condition, investment attractiveness assessment, evaluation of financial and economic viability of the counterparties, etc. The proposed tools are beyond the scope of the above problems and can be used in other dynamic problems with incomplete data for modeling economic objects.

Cited by 0 works.

Authors

References