Modelling Dynamic System for Collective Learning Behaviors

In this paper, we will look into a typical discrete dynamic model of common types of collective behaviors. Collective learning behavior of human group with interactions is first modelled according to the classical herding theory in psychology, and then the model is constructed and simulated based on the cellular automaton (CA) method. In our method, we assume that individuals learn to make their decisions repeatedly, and each individual´s behavior at coming time step is driven by situation of system and by individual´s anticipation with respect to future decisions of other individuals at current time step. CA is employed to compute and observe the visualized long-term behaviors of the system at higher group level, and a set of important statistical characters of the system are discussed. Our method is helpful for researches of modelling and simulating social dynamic systems and of researching higher-level behaviors of collective system with learning techniques.