Summary

In this chapter, we introduced you to a number of clustering algorithms, including K-means and AHC. We also introduced you to a variant of AHC that leverages a spatial constraint via the spatial weights matrix to develop geographically constrained clusters known as regions.

For each of the clustering models, we evaluated the clusters through cluster profiling. We produced maps of each cluster and also calculated a variety of descriptive statistics, including cluster tract counts, average cluster tract area, and mean values. We then used this information to produce choropleth maps of each of the clustering algorithms.

In the final section of the chapter, we introduced you to the Calinski-Harabasz score, the Davies-Bouldin score, and the Silhouette score, which are common mathematical measures of clustering performance. Even though the K-means-based model scored the best mathematically, it may not be the clustering model that makes the most sense for your use case. For...