MODERN PROBLEMS OF RESEARCH AND APPLICATION OF MULTI-AGENT SYSTEMS IN ASSESSMENT OF COLLECTIVE ACTIVITY

Abstract

Currently, the most widely used approach to modeling complex systems are models built using multi-agent systems that, the models play a major role in the field of Artificial Intelligence. Agents are the main element in these models. The order of solving problems is distributed according to a set of rules between agents included in a certain team. Agent systems have properties that allow learning such phenomena as self-organization and collective intelligence. These features allow using MultiAgent Systems with equal success in both scientific-theoretical and practical-applied research. With the help of MAS, it is possible to find simpler and more uniform solutions for solving complex problems. Multi-agent systems (MAS) are an interesting area of research in artificial intelligence and computer science. They involve numerous autonomous agents interacting in an environment to achieve individual or collective goals. The article analyzes scientific works devoted to scientifictheoretical and technical-practical problems and solved with the help of MAS. The shortcomings of
multi-agent systems are investigated and solutions are proposed. Based on the results obtained, a
scientifically substantiated interpretation of modern and traditional methods is given. It is shown that
mainly scientific works offer methods for making decisions in the direction of modern scientifictechnical problems. The article also analyzes research works devoted to the application of IT methods and tools using MAS, and studies control and forecasting models using neural networks and intellectual methods