We study the DeGroot model for continuous opinion dynamics under the influence of innovations. In the original model, individuals＇ opinions, after given their initial values, evolve merely according to the given lear...We study the DeGroot model for continuous opinion dynamics under the influence of innovations. In the original model, individuals＇ opinions, after given their initial values, evolve merely according to the given learning topology. The main contribution of this paper is that external innovation effects are introduced： each individual is given the opportunity to change her opinion to a randomly selected opinion according to a given distribution on the opinion space and then the external opinion is either adapted by the individual, or combined into her learning process. It turns out that all the classical results of the DeGroot model are violated in this new model. We prove that convergence can still be guaranteed in the expectation sense, regardless of the learning topology. We also study the steady distributions of opinions among the society and the time spent to reach a steady state by means of Monte-Carlo simulations.展开更多
基金Supported by the 973 Program(2010CB731405)National Natural Science Foundation of China(71101140,11471326)
文摘We study the DeGroot model for continuous opinion dynamics under the influence of innovations. In the original model, individuals＇ opinions, after given their initial values, evolve merely according to the given learning topology. The main contribution of this paper is that external innovation effects are introduced： each individual is given the opportunity to change her opinion to a randomly selected opinion according to a given distribution on the opinion space and then the external opinion is either adapted by the individual, or combined into her learning process. It turns out that all the classical results of the DeGroot model are violated in this new model. We prove that convergence can still be guaranteed in the expectation sense, regardless of the learning topology. We also study the steady distributions of opinions among the society and the time spent to reach a steady state by means of Monte-Carlo simulations.
