I have developed an IP clustering program much like the blackhat USA application shown here: https://www.blackhat.com/docs/us-16/materials/us-16-Wolff-Applied-Machine-Learning-For-Data-Exfil-And-Other-Fun-Topics.pdf
However my application has the option of using nessus exported files as well as nmap whilst concatenating the features together, this process adds the feature of vulnerabilities into the clustering algorithm.
The end result outputs graphs and statistics whilst showing the most vulnerable machines on the network. This is all done automatically in a few minutes without the user having to manually look through large files on networks with hundreds to thousands of IP's.
unfortunately I cannot provide full examples as of yet as its for my thesis, yet to be submitted.
I was wondering if this application would be useful for any network admins or pentesters, id love to give back to the community. In which case ill release it open source.