A method for inductive estimation of public transport traffic using spatial network characteristics

An inductive method based on spatial network characteristics and geographical neighbourhood is proposed, which allows to extrapolate a sample of representative passenger counts for an existent public transport network. The most important predictors can directly be deduced from a logical model of the spatial network. Furthermore, the attractiveness of stops can be modelled by spatial densities in their geographical neighbourhood. The method is formally specified using a typed logic.