Abstract

Automatic Region Building for Spatial Analysis
Track: Education
Authors: Diansheng Guo, Hu Wang

High-resolution spatial data become increasingly available but it is often inappropriate to use the default geographic units to perform spatial analysis due to unstable estimates with small areas (e.g. cancer rates for census tracts). Regionalization is to aggregate small units into larger areas while optimizing a homogeneity measure. For spatial analysis, regionalization may help remove spurious data variation and discover hidden patterns in data. This research introduces several improvements to a recent group of regionalization methods and conducts evaluation experiments with synthetic data sets to assess the capability of different regionalization methods. One of the major improvements is the integration of a local empirical Bayes smoother (EBS) with the regionalization methods. Evaluation results show that the new methods (integrated with EBS) perform significantly better than their original versions and other methods (including the EBS method on its own) in terms of detecting the true patterns in the synthetic data sets.