Animations to accompany the following manuscript:
Schafer, T. L., Wikle, C. K., & Hooten, M. B. (2020). Bayesian Inverse Reinforcement Learning for Collective Animal Movement. arXiv preprint arXiv:2009.04003.
The following animations show the observed trajectories for one guppy experiment and two simulations. Our modeling framework did not include a component for step length so the simulations assumed a constant step length throughout.
Observed trajectories for one all male experiment
Simulated state transitions using probabilitistic transitions induced by the posterior mean of the costs-to-go with initial positions from the observed experiment
Simulated state transitions choosing the least costly state as estimated by the posterior mean of the costs-to-go with initial positions from the observed experiment
