Buses of Cuernavaca - an agent-based model for universal random matrix behavior minimizing mutual information

摘要

The public transportation system of Cuernavaca, Mexico, exhibits randommatrix theory statistics [1]. In particular, the fluctuation of times betweenthe arrival of buses on a given bus stop, follows the Wigner surmise for theGaussian Unitary Ensemble. To model this, we propose an agent-based approach inwhich each bus driver tries to optimize his arrival time to the next stop withrespect to an estimated arrival time of his predecessor. We choose a particularform of the associated utility function and recover the appropriatedistribution in numerical experiments for a certain value of the only parameterof the model. We then investigate whether this value of the parameter isotherwise distinguished within an information theoretic approach and givenumerical evidence that indeed it is associated with a minimum of averagedpairwise mutual information.