Neighboring index between clusters on a map
We have a situation that can be simply put as follows.
We have many points on a map. We built a network model by connecting the points that are closeby on the map. The points are clustered based on an independent set of features. Let’s say the clusters are pretty well bundled on the map. Do you know of any metric or index indicating how each cluster is connected with other clusters? Naively, between cluster 1 and 2, I can count the number of points in cluster 1 that are connected to cluster 2 and vice versa. Then, we can normalize the counts by the total number of counts in the two clusters. However, I am still thinking that there should be a more intelligent way to do it. Could you please point me where I need to check?
• 13 views
Read more here: Source link