Kernel density analysis is a type of cluster analysis. It helps us to detect areas where there are increased intensity of a phenomenon. This link provides instructions on how to calculate kernel densities. Usually, the main issue is with the search radius. Often the default search radius does not give recognizable clusters, so you may have to increase the search radius. Feel free to play around with the search radius until your clusters look smooth and realistic. The image below provides an idea of what clusters looks like after running a kernel density or hot spot analysis.
No comments:
Post a Comment