Artificial intelligence methods and models of reflexive behavioral economics
Annotation. Solving the problems of managing the behavior of economic agents in modern socio-economic systems requires the use of non-standard approaches and tools of economic and mathematical modeling. The possibilities of linking research into the interactions of economic agents within the framework of the reflexive approach and the theory of behavioral economics with the capabilities of formal methods and models of artificial intelligence for studying the patterns and mechanisms of decision-making by economic agents and solving the problems of forecasting and managing their behavior to ensure the achievement of the goals of socio-economic systems are considered.
The study of behavioral economics, based on the theory of reflexive management, is the mainstream of modern economic science [1-5]. The reflective approach is based on the study of decision-making procedures based on the cognitive mechanism of perception and interpretation of information by economic agents. However, the task of diagnosing the behavior of economic agents for the purpose of further management to obtain the desired results of their activities or decision-making in socio-economic systems is extremely difficult due to weak formalization, uncertainty of agent behavior, rapid changes in the modern information space, and the subjectivity of decision-making processes. Therefore, solving the problems of managing the behavior of economic agents in socio-economic systems requires the use of non-standard approaches and modern tools of economic and mathematical modeling.
The hypothesis of the study is that on the basis of a reflexive approach using artificial intelligence methods and models, it is possible to identify the features of behavioral manifestations and use it in managing of economic agents to increase the efficiency of the functioning of socio-economic systems and direct the actions of economic agents in the direction of public interests. To prove the hypothesis put forward, it is proposed to link research into the interactions of economic agents within the framework of a reflexive approach and the theory of behavioral economics with the capabilities of formal methods and models of artificial intelligence for studying the patterns and mechanisms of decision-making by economic agents for their further use in solving problems of forecasting and managing their behavior to ensure the achievement of the goals of socio-economic systems.
Given the specifics of the research, the definition of artificial intelligence is considered from the standpoint of its classical understanding. Thus, it is a mathematical and software toolkit, the principles of which reproduce the processes of human decision-making (according to biological “bottom-up” or semiotic “top-down” approaches), primarily for solving cognitive tasks characteristic of humans[2]. The methodological basis for the development of conceptual provisions and a set of relevant economic and mathematical models for managing the behavior of economic agents in socio-economic systems is the reflexive approach and the theory of reflexive management (in particular, the nudge theory of R. Thaler [1]). This approach considers the decision-maker as an economic agent with a complex structure of ideas, on the basis of which he makes decisions [5]. The structure of ideas is determined by the components of the decision-making mechanism – the reflexive characteristics of economic agents, on which the result of their decision-making directly depends, and as a result, the effectiveness of the functioning of socio-economic systems where economic agents interact [6].
The proposed integrated scientific concept defines the key parameters of managing the behavior of economic agents in socio-economic systems within the framework of evaluating the reflexive characteristics of agents that determine the specifics of their behavior. The corresponding conceptual provisions provide for the construction and effective use of appropriate methods and models of behavior diagnostics and reflexive management of the behavior of economic agents in socio-economic systems using artificial intelligence methods and models.Methods and models for diagnosing the behavior of economic agents involve the use of neuro-fuzzy analysis of the behavior of economic agents (semiotic approach), in particular, the use of fuzzy logic methods for processing the results of questionnaires of economic agents or information from open or specialized digital databases and further neural network modeling based on of Kohonenself-organizing maps(biological approach) to identify the specific manifestations of the behavior of economic agents in socio-economic systems[6].In turn, the proposed reflexive models of behavior description allow predicting the results of the decision-making process by economic agents in socio-economic systems. Further management of the behavior of economic agents based on the results of forecasting using methods of information and reflexive influence will ensure increased efficiency in achieving the defined goals of socio-economic systems.
Possible benefits of managing the behavior of economic agents in socio-economic systems within the framework of the proposed conceptual provisions may be:
- taking into account the subjective features of the decision-making process by economic agents to increase the efficiency of managing socio-economic systems;
- increasing the efficiency of information interaction between economic agents and the quality of decisions in accordance with the goals of socio-economic systems;
- directing the actions of economic agents towards economic development and increasing the efficiency of the functioning of socio-economic systems at the macro and micro levels;
- increasing the effectiveness of actions of economic agents in socio-economic systems due to their direction in the direction of public interests and social and economic development;
- ensuring information security of economic agents in the process of interactions in socio-economic systems.
Thus, the proposed conceptual provisions based on a reflexive approach and methods and models of artificial intelligence form the scientific basis for managing the behavior of economic agents to ensure controllability and increase the efficiency of the functioning of socio-economic systems at macro and micro levels by ensuring the direction of the actions of economic agents in the direction of public interests and economic development.
Література
- Thaler R. H. Misbehaving: The Making of Behavioral Economics. N.-Y. : W.W. Norton & Company, 2015. 432 p.
- Turlakova S., Shumilo Ya., Lohvinenko B. The Future of Behavioral Economics: AI Tools in the Digital Space Eds. S. Eds. S.Turlakova; National Academy of Sciences of Ukraine, Institute of Industrial Economics of the NAS of Ukraine. Kyiv: Akademperiodyka, 2024. 170 p.
- Turlakova S., Lohvinenko B. Artificial Intelligence Tools for Managing the Behavior Economic Agents at Micro Level. Neuro-Fuzzy Modeling Techniques in Economics. 2023, 12. P. 3-30. http://doi.org/10.33111/nfmte.2023.003.
- Turlakova S. S., Shumilo Ya. M. Influence of AI Tools on Consumer Behavior Management in Digital Marketing. Science and Innovation. 2025, 21(1). Р. 67–81. https://doi.org/10.15407/scine21.01.067.
- Турлакова С. С. Рефлексивне управління поведінкою економічних агентів у цифровому просторі із використанням інструментів штучного інтелекту. Актуальні проблеми системного аналізу та моделювання процесів управління / За ред. В. Пономаренка, Л. Гур’янової, Я. Пеліової, Е. Ніжинського.Братислава-Харків : ВШЕМ–ХНЕУ ім. С. Кузнеця, 2023. С. 268-283.
- Turlakova S. Modeling the values of reflexive characteristics of agents within the management of herd behavior at the enterprises. Neuro-Fuzzy Modeling Techniques in Economics. 2022. №11. P. 48-71. http://doi.org/10.33111/nfmte.2022.048.