Abstract
The contrast between entropy measured in thermodynamics and information theory is discussed in terms of its implications for trip distribution and related urban-modelling procedures based on entropy maximisation or information minimisation. Models of spatial interaction using α = 2 entropy measures are shown to be improper because of the limited properties of these measures. A dynamic entropy-maximising Markovian framework for projecting trip and location distributions is suggested in conjunction with an earlier proposal of a ‘Bayes chain’ based on α = 1 information minimisation.

This publication has 14 references indexed in Scilit: