A Reinforcement-Learning-Based Asynchronous Group Decision Making System

Group decision making is a typical representative of the group tasks. How to define the weight of decision makers (DMs) and how to adjust their decision preferences are two important research issues in group decision making. In order to solve the above problems, this paper proposes a learning-exploration-integration-feedback-consensus (LEIFC) group decision making system. The LEIFC system includes three processes: DMs selection, decision making and consensus adjustment. In the DMs selection process, based on the reinforcement learning, the system selects a subset from the participants to form a DMs set. In the decision making process, the system promotes the undecided to make decisions through an asynchronous method. In the consensus adjustment process, an auto adjustment consensus mechanism based on particle swarm optimization algorithm is applied to adjust the decision preferences of DMs. The numerical simulations show that on the basis of selecting the DMs with the optimal performances, the LEIFC system can make the group decision making results more accurate.