Stratifying and predicting patterns of neighbourhood change and gentrification ‐ an urban analytics approach

While recent debates have widely acknowledged gentrification’s varied manifestations, success in enumerating and disentangling the process and its defining features from other forms of neighbourhood change at-scale and across entire cities, has remained largely elusive. This paper addresses this gap and employs a novel, open and reproducible urban analytics approach to systematically examine the past and future trajectories of neighbourhood change using London, England, as a case-study example. Using suites of datasets relating to population, house prices and built environment development, the nature of gentrification’s mutations and its spatial patterns are extracted through a multi-stage data dimensionality reduction and classification methodology. Machine Learning is subsequently adopted to model gentrification’s observed trends and predict its future frontiers with interactive visualisation methods offering new insights into gentrification’s projected dynamics and geographies