Принятие управленческих решений в кризисных ситуациях на основе нейронной сети «дерево решений»
Сенин А.С.1, Лясников Н.В.2,1
1 Российская академия народного хозяйства и государственной службы при Президенте Российской Федерации
2 Институт проблем рынка РАН
Download PDF | Downloads: 32 | Citations: 8
Journal paper
Economics and society: contemporary models of development (РИНЦ)
опубликовать статью
Volume 9, Number 1 (January-March 2019)
Indexed in Russian Science Citation Index: https://elibrary.ru/item.asp?id=38479966
Cited: 8 by 07.12.2023
Abstract:
Purpose: based on the analysis of literature sources, as well as international experience in the development of virtual situation center technology to explore the features of the neural network "decision tree" to support management decision-making in crisis situations. Materials and methods: the methodological basis of this article is the literature on the development of virtual situation center technologies, as well as open analytical materials on the experience of using the neural network "decision tree" to support management decision-making in crisis situations. Results: in the present article the essence of virtual situation center technology is defined: and its role to support management decision – making in crisis situations is analyzed the use of neural network "decision tree" to support management decision-making in crisis situations. Conclusions: algorithms of machine support of management decisions in critical situations can be used in many areas, as a rule, where it is necessary to automate complex tasks, for which it is customary to use the knowledge and experience of a person. Today, neural networks are widely used. Neural networks are based on the idea that a neuron is a simple element that can be modeled. And all the complexity of human thinking comes from a huge number of neurons (in the human brain there are more than 1010) and the complexity of the links between them. Thus, for example, in banking structures to support management decision-making, a neural network can detect complex, nonlinear and non-trivial relationships between the characteristics of the client and his solvency, whether he will return the loan on time or not, which can not be detected by logistic regression and classification trees. Today, neural networks are widely used in large financial companies. So, for example, Lloyds Bowmaker Motor Finance uses neural networks "decision tree" to make decisions in car loans; security Pacific Bank – in lending to small businesses and so on. However, the most important drawback of neural network models is the complexity of interpretation, as the structure of the neural network does not allow to describe the relationship in a simple way. Application. The findings and results of the study can be used in the further use of the neural network "decision tree" to support management decision-making in crisis situations.
Keywords: latest technologies, virtual center, neural networks, decision tree management decisions, crisis situations, computer programs, artificial intelligence
References:
2. Shevelev O.G. Tyuterev V.V. Multi-layer perceptrons in the problem of resolving controversial authorship of the text. In: Collection of works of the scientific and technical conference "Neuroinformatics-2003", January 29-31, 2003, Moscow. M.: MEPI, 2003. Part 2. P. 206–212 (in Russ.)
3. Nikishov S.I. Application of fuzzy composition for logistics flows modeling. Russian Journal of Innovation Economics. 2017; 7(3). doi: 10.18334/ vinec.7.3.38197 (in Russ.)
4. Tyuterev V.V., Novoseltsev V.B. Theory of neural networks. Tomsk: Publishing House of Tomsk State University, 2000. 56 p. (in Russ.)
5. Luizova L.A. From setting the problem to making a decision. Petrozavodsk: Publishing House of PetrSU, 2002. (in Russ.)
6. Kharasova A., Blazhenkova N. Managerial Decision-making Technology in Theory and Practice of Management. Leadership and Management. 2015; 2(2). doi: 10.18334/lim.2.2.598 (in Russ.)
7. Tyunik O., Pankina T. Decision-making in the economic activity of trade organizations in terms of risk. Russian Journal of Retail Management. 2015; 2(3). doi: 10.18334/tezh.2.3.592 (in Russ.)
8. Wassermen F. Neurocomputer technology. M.: Mir, 1992 (in Russ.)
9. Short S. Neural networks: an algorithm of back propagation [Electronic resource]. Access mode: http://www.orc.ru/ stasson/n2.zip (in Russ.)
10. Yezhov A.A., Shumsky S.A. Neurocomputing and its applications in economics and business [Electronic resource]. Access mode: http://soft. neurok.ru/pub/lectures.shtml, free (in Russ.)
11. Shevelev O.G., Petrakov A.V. Classification of texts with the help of decision trees and neural networks of direct propagation. TSU Bulletin. Tomsk: Vol. University, 2005 (in Russ.)
12. Gluschenko V.V., Gluschenko I.I. Theoretical aspects of preventive crisis management in corporations. Leadership and Management. 2017; 4(4). doi: 10.18334/lim.4.4.38373 (in Russ.)
13. Marsden T., Sonnino R. Rural development and the regional state: Denying multifunctional agriculture in the UK. Journal of Rural Studies. 2008; 24(4): 422–431.
14. Reshetov K.Yu., Khoroshavina N.S., Mysachenko V.I., Komarov V.Yu., Timofeev M.I. State policy in the area of implementation of innovations in industrial production. The Journal of Social Sciences Research. 2018; (S3): 271–276.
15. Cheyns E., Daviron B., Djama M., Fouilleux E., Guéneau S. The standardization of sustainable development through the insertion of agricultural global value chains into international markets. Sustainable Development and Tropical Agri-chains. Springer, Dordrecht, 2017. P. 283– 303.
16. Giannakis E., Bruggeman A. The highly variable economic performance of European agriculture. Land Use Policy. 2015; 45: 26–35.
Страница обновлена: 19.05.2025 в 08:07:05