Jūs esate čia: Pradžia Summary

Vaizdai iš knygos (paspaudus, atidaromas teksto pavyzdys)

Banner

Kitos VVA nuorodos

Banner
Banner
Banner
Banner
Banner
Banner
Banner
Monografija: summary PDF Print E-mail
Written by Administratorius   
Friday, 02 April 2010 14:51

Monograph

Computational Intelligence and Multi-Agent Systems in the Social Domain: Methods, Applications, and Perspectives

D. Plikynas and P. Daniušis

Recent turbulent years of financial meltdown and economic recession have witnessed an explosion of interest and research activity in the area of complexity economics, which is related to evolving social networks and the knowledge economy. Many researchers are looking for alternative ways to model and simulate emergent social phenomena.

Currently applied traditional top-down modeling approaches mostly do not satisfy modern social research, which is still fragmented by a lack of cohesive modeling media for understanding and simulating the highly complex nature of social phenomena. There is a clearly observable need to design bottom-up models which can self-organize, adapt, and evolve like living systems. In reality, standard models fail to describe such an evolving complexity because they (i) do not have the same order of complexity (or degrees of freedom) and (ii) do not evolve together with the object of investigation. Today, such properties are mostly simulated by using intelligent agents and multi-agent systems.

This monograph not only systematically presents currently available alternative approaches in the area of computational intelligence methods but also gives some clues about how these methods can be naturally integrated into the construction of intelligent agents as well as artificial societies. The book starts with the construction of intelligent agents and ends with the multi-agent systems which are used for modeling and simulating modern social phenomena.

This highly interdisciplinary monograph discusses some new insights into the fundamental properties of information-rich social networks. The study of coordinated models goes beyond computer science in that evolutionary computation, behavioral sciences, social sciences, business management, artificial intelligence, and logistics also somewhat strictly deal with how social agents can properly coordinate with each other and emerge as globally coherent behaviors from local interactions. Therefore, the major question we address in this book is: what is the natural and most efficient way of simulating complex mechanisms for the networking and interaction of information in order to reflect the observed multiplicity of social behavior?

The book has three interconnected parts. The first part starts with the methodical approaches and concepts concerning computational intelligence paradigms like evolutionary computation, neural networks, fuzzy logic, hybrid systems, and principal component analyses. Then the monograph explores multi-agent platforms and agent-based computational models.

The second part proceeds with applications of computational intelligence methods and multi-agent platforms in the social domain. The main emphasis is on the interdisciplinary co-integration of both computational intelligence and multi-agent platforms, as both synergistically supplement each other. The style and method of the ideas presented are a mix of traditional economics (and finance), econophysics, information theory, computational intelligence, and multi-agent systems. The original applied research cases are taken from articles previously accepted by journals devoted to these fields.

Finally, the third part is concerned with the design concepts and main principles for the field-based simulation of modern information societies. A number of broader theoretical questions and even some controversial philosophical ones are raised. This part outlines some perspectives, ending up with a unique social simulation paradigm which seeks to shed new light on a conceptually new way of simulating the complex social processes taking place in an information economy. In brief, this study proposes a virtual information field-based communication model in which communicational processes are modeled on a virtual computational grid.

Last Updated on Tuesday, 06 April 2010 13:45