Dynamical networks as a tool for system analysis and exploration

Abstract

This paper attempts to bring together methods and tools across three selected areas of Computational Intelligence pertaining to complex systems: dynamical networks, statistical machine learning, and multi-agent technologies. The paper begins with a discussion of computational challenges arising from growing complexity in modern operating environments. Then the discussion proceeds to an overview of intelligent tools that can be used to address these challenges. The emphasis is given to dynamical networks as the means for consolidating ideas and approaches. An illustrative example shows how the network-based tools can be used to model complex socio-technical systems, to identify hidden interdependencies among their components, and to explore their vulnerabilities in simulations.