Building size modelization

ABSTRACT New challenges in the efficient management of cities depend on a deep knowledge of their inner structures. It is therefore very important to have access to reliable models of cities characteristics and organization. This paper aims at providing and validating a stochastic modelization based on statistical data of buildings parameters which can be useful as an entry for many other models considered in a wide range of fields where buildings structure is a main factor of a thorough modelization of cities. The interest of such an approach is highlighted through the detection of errors in the data or as a tool for visual clustering.