Classification of Residential Areas in the Three Largest Dutch Cities Using Multidimensional Data

Abstract

This article explores the variety of residential area types in the three largest Dutch cities: Amsterdam, Rotterdam and The Hague. The data are characterised as multidimensional, as they pertain to socio-demographic and other `objective' indicators produced by official statistics, as well as to more `subjective' location preferences expressed by the residents themselves. The method used is a clustering and classification approach based on two-dimensional models generated by the self-organising map (SOM) and learning vector quantisation (LVQ). Using sub-district-level data, the study shows that, while some characteristic residential milieus can be found in all three cities, they have a broader variety in Amsterdam than in Rotterdam or The Hague. This supports the commonly held presumption that Amsterdam is less segregated than Rotterdam or The Hague in terms of intraurban housing patterns.