|
Abstract:
|
Challenges occur while modeling complex collective animal movement. Particularly, the modeling framework needs to not only capture the inherent nonlinear interaction behaviour between animals presented in collective movement processes, but also should account for uncertainty in data, model, and parameters. The self-propelled particle (SPP) model from Vicsek et al. (1995) is a widely applied agent-based model for modeling the collective behaviour of objects. Here, we propose a hierarchical Bayesian framework including a modified SPP model to describe collective motion of objects and a Von Mises distribution to account for uncertainty in animal movement. We illustrate the hierarchical Bayesian SPP methodology with a simulation study and by applying to the movement of guppies.
|
