I have a non-directed graph that represents the connectivity between regions of a map. I'd like to identify groups of nodes (regions) that could be removed without creating graph partitions.
What I have tried:
Walking the tree (BFS, DFS...), storing the depths and selecting the nodes with the higher depth (O(n)). Once calculated, I can update the depths in O(~1) on each removal-addition by checking the depth of neighbour nodes (connectivity does not exceed a certain threshold)
Is there a cheaper way to do this? Also finding graph literature is also very hard if you don't know the academical term for the problem. My graphs are between 200 and 500 nodes.