Filippo Simini, Amos Maritan, Zoltán Néda
Models of human mobility are derived within the framework of the recently introduced radiation model. A unified continuous approach is used for computing the probability to observe a trip to any region. Fluxes among all areas, defined by a generic spatial subdivision, are computed and are shown to lead to previously proposed models, like the intervening opportunities model, or new models, like the high expectations model, depending on the benefit distribution. Comparison with observational data offered by commuting trips extracted from the US census data set shows the validity of such models. Finally, we discuss the advantages of the continuum approach, we illustrate the fluxes' additivity property, and we use our theoretical framework to derive some special cases of gravity models from first principles. These findings suggest that the complex topological features observed in large mobility and transportation networks may be the result of a simple stochastic process taking place on an inhomogeneous landscape.
View original:
http://arxiv.org/abs/1206.4359
No comments:
Post a Comment