Institute of Information Theory and Automation

You are here

Normative theory and algorithms of distributed decision making under uncertainty and incomplete knowledge

Type of Work: 
ÚTIA AV ČR, v.v.i., AS department, 266052274
Adaptive systems, distributed systems, advisory systems, Bayesian learning, probabilistic design, probabilistic coopeartion

Real-life processes arise as a result of relatively independently deciding, mutually influencing parts, which are modelled as multi-agent systems. A complete prescriptive algorithmically feasible theory sutitable to solutions of specific problems does not exist. Preliminary results indicate that such a theory can be created by combining Bayesian decision making, dynamic mixture modelling and fully probabilistic design of decision strategies. This emergent area having applications, for instance, in technology or electronic democracy opens a range of theoretical, algorithmic, experimental and software problems, which suit as topics of 2-3 Phd theses.    


It can be studied at FJFI, FEL, ZČU or according to a specific agreement

[1] Andrýsek J., Kárný M., Kracík J. (Eds.): Multiple Participant Decision Making. (International Series on Advanced Intelligence. 9). Advanced Knowledge International, Adelaide 2004, 182 pp. [2] Kárný M., Guy T. V.: On dynamic decision-making scenarios with multiple participants. In: Multiple Participant Decision Making. (Andrýsek J., Kárný M., Kracík J. eds.). (International Series on Advanced Intelligence. 9). Advanced Knowledge International, Adelaide 2004, pp. 17-28. [3] Kárný Miroslav, Guy Tatiana Valentine, Bodini A., Ruggeri F. : Cooperation via sharing of probabilistic information , International Journal of Computational Intelligence Studies, p. 139-162 [2009] [4] Kárný Miroslav, Herzallah R. : Scalable Harmonization of Complex Networks With Local Adaptive Controllers , IEEE Transactions on Systems Man Cybernetics-Systems vol.47, 3 (2017), p. 394-404

2018-08-13 09:31