AS seminář: Tools for Sharing of Decision Elements Among Imperfect Bayesian Participants

Bayesian decision theory provides a strong theoretical basis for a single-participant decision making under uncertainty, that can be extended to multi-participant problems. However, Bayesian decision theory assumes unlimited abilities of a participant to probabilistically model participant's environment and to optimise decision-making strategy. The lecture describes results presented at NIPS 2010 workshop "Decision Making with Imperfect Decision Makers". A methodology is outlined for sharing of knowledge and preferences among participants, that helps to overcome the non-realistic assumption on participants' unlimited abilities.