Social Learning and Collective Analysis in Boolean Networks

• 作 者: LouYoucheng;HongYiguang;
• 机 构: Key Laboratory of Systems and Control,Academy of Mathematics and Systems Science Chinese Academy of
• 出 处: 2010第六届全国多智能体系统与控制学术年会
• 发表日期: 2010年期

　　In this paper,the learning behavior of Boolean networks is investigated.Each agent starts with a prior belief about the true state.Next,each agent randomly selects an action with probability of prior belief,observes the actions of neighbors,and makes decision between two possible actions based on majority principle.In the decision process,if all agents dont receive private signal about the underlying state of the world,then we show that all agents will make the same decision under mild assumptions and show that the common decision is false with positive probability.However,if at least an agent receives private signal,then update their private belief in Bayesian way,we can show that each agent will make the same decision and the common decision is right with probability one,that is,the asymptotic learning is achieved almost surely.