Dynamic neural network models of bankruptcy of corporations with incomplete data
Please sign in so that we can notify you about a reply
The monograph is devoted to a complex and practically unexplored problem of neural network modeling of the development of bankruptcies of corporations in dynamics. The complexity of these models follows from the specific incompleteness of data caused by legal reasons, and a strong noise of data. A method of optimizing the structure of the neural network is proposed in combination with its Bayesian regulator, as well as an algorithm for compression of variables based on the generalized function of Harrington"s desirability. The conceptual basis of neural network modeling and the neurosetic logistic dynamic method, which restores incomplete data during the solution of the problem of approximation of the dependence “entrance-output”, has been developed on the basis of general system laws. For the first time, hybrid neural network models of illegal bankruptcy of legal entities were considered. The theoretical ideas put forward in detail are illustrated by applied tasks and justified by computational experiments on real data. The monograph material is 90% original, summarizes and develops methods of neural network modeling of bankruptcies from previous books of the authors.
For students, undergraduates and teachers of a wide range of universities, as well as scientists interested in the problems of neural network modeling in the field of financial management and economic security of enterprises
For students, undergraduates and teachers of a wide range of universities, as well as scientists interested in the problems of neural network modeling in the field of financial management and economic security of enterprises
Author:
Author:Gorbatkov Stanislav Anatolyevich
Cover:
Cover:Hard
Category:
- Category:Business & Money
ISBN:
ISBN:978-5-907244-86-3
No reviews found