With the advent of extremely affordable computing power, the world is becoming filled with distributed systems of computationally sophisticated components. However, no current scientific discipline offers a thorough understanding of the relation of such "collectives" and how well they meet performance criteria.
"Collectives and Design of Complex Systems" lays the foundation for the study of collective intelligence and how these entities can be developed to yield optimal performance. Partone describes how some information-processing problems can only be solved by the joint actions of large communities of computers, each running their own complex, decentralized machine-learning algorithms. Part two offers general analysis on the dynamics and structures of collectives. Finally, part three addresses economic,
model-free, and control-theory approaches to designing these complex systems. The work assumes a modest understanding of basic statistics and calculus.
Using an approach that integrates key theoretical principles with applications in real-world scenarios, this unique monograph surveys the latest research on the dynamics of collectives, their A.I. aspects, and critical design issues pertaining to them. Computer scientists, computer engineers, and all practitioners, researchers, and graduate students with an interest in this new and growing field will find the book an authoritative introduction and resource.
