Modelling Uncertainty-exposed Team Decision-making in Multi-agent System

摘要

Organizational systems work in terms of uncertainty in uncoordinated decision-making. Information flows in systems and environments are not always sufficient to lead to an exact probable outcome. In this connection, there is a need in mathematical modelling of team decision-making as a process to overcome information uncertainty in a formalized way. To do this, there is a proposal to use the multi-agent model based on the neural model structure. Authors have given a systematic presentation of important provisions from the theory of optimal control over systems, including decision-making in two-alternative and multi-alternative cases. The reviewed program model confirms convergence of the game neuroagent method for decision-making. Effectiveness of the method was estimated using characteristic functions of average gains and errors in team choices of the best solutions in decision-making. Reliability of findings is confirmed by repeated values of method characteristics calculated for various sequences of random variables.