Multi-agent Technologies

About the course

A significant reduction in the size of computing devices and an increase in their speed provide new opportunities for solving traditionally complex, multidimensional problems of optimization, resource allocation, logistics, etc. Multi-agent technologies (MT) were born at the junction of modern distributed methods for solving various problems, parallel programming and artificial intelligence. At the heart of MT is a decentralized approach to solving problems, in which dynamically updated information in a distributed network of intelligent agents is processed not in some center, but directly at the agents along with locally available information from neighbors. At the same time, both the resource and time costs for communication in the network and the time for processing and decision-making in the center of the entire system (if there is one) are significantly reduced.

Emergent intelligence (intellectual resonance, swarm intelligence) is an occurrence of unexpected properties that a system possesses, but none of its individual elements has.
Key feature is the dynamics and unpredictability of the decision making process.

Enroll in a course:
Course's authors:
©Санкт-Петербургский государственный университет
2021 год