Averager-copier-voter models for hybrid opinion dynamics in complex networks
Provides a theoretical framework for understanding opinion formation in heterogeneous multi-agent systems, relevant to social networks and distributed decision-making.
The paper introduces a hybrid model for opinion dynamics combining continuous averaging agents and discrete copying/voting agents, proving consensus under ergodic network conditions and characterizing convergence time in terms of network structure and agent distribution.
A hybrid model for opinion dynamics in complex multi-agent networks is introduced, wherein some continuous-valued agents average neighbors' opinions to update their own, while other discrete-valued agents use stochastic copying and voting protocols. A statistical and graph-theoretic analysis of the model is undertaken, and consensus is shown to be achieved whenever the network matrix is ergodic. Also, the time required for consensus is characterized, in terms of the network's graph and the distribution of agents of different types.