Decision making (DM) is an essential component of the problem-solution at many levels - from decisions at international level (e.g. aimed at securing nuclear safety or preventing flood damage), to individual organisations (e.g. increasing productivity or improving urban traffic), systems (e.g. process or robot control, fault detection, medical diagnostics) and to the level of individual human beings following multiple personal aims in their changing environments.

Any man-made complex system is composed of DM units called participants. Participants can he machines, groups of humans or their combinations. Attempts to optimise centrally the overall performance of a collection of mutually interacting participants soon reach complexity barrier that allows performance improvements only at unacceptable costs. Use of sophisticated distributed or multiple-participant DM methodologies is then an only viable way towards desirable high efficiency. Excellent particular variants exist that overcome the complexity barrier by exploiting specificity of their application domains. None of them is yet able to serve as a common domain-independent pattern and a real need for theory of multiple-participant DM persists.

Book Contents

This book brings together contributions of experts of different backgrounds who inspect various aspects of the problem, push the state of the knowledge towards the dreamt-of theory and open a range of questions to be addressed. At least the last item makes this collection worth of reading.

About the Author

Editors represent two generation of researchers of Department of Adaptive Systems of the Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic. All of them got their degrees Ing. (MSc) from Czech Technical University in Prague. Josef Andry'sek graduated in Software Engineering. His research is focused on recursive estimation of high-dimensional finite probabilistic mixtures that serve as universal approximation of non-linear stochastic systems. 

Miroslav Karny graduated in Theoretical Cybernetics and received his CSc (PhD) and DrSc degrees from Czechoslovak Academy of Sciences both in Technical Cybernetics. His research interests cover various theoretical, algorithmic and application aspects of dynamic decision-making under uncertainty. Adaptive advising and control based on recursively estimated finite dynamic probabilistic mixtures and their fully probabilistic optimisation dominate his current research. 

Jan Kracik graduated in Mathematical Modelling. His research on fair governing led him to inspection of combining knowledge and aims in multiple-participant decision-making.