Estimating swarm parameters by evolutionary learning.
Richard D. Coates, Simon Hickinbotham · 2011
If we assume that the collective dynamics of wild animals can be modelled, it would be desirable to recover the dynamics of the model via interaction with them. In this paper, we demonstrate that it is possible to recover the parameters of a shoaling model used by a swarm. This can be achieved by evolving the parameters of a single agent that interacts with the swarm. We present an evaluation of this approach, using a genetic algorithm to learn the parameters of a shoaling model.