4 ms·
Show HN: KarateClub a Python library for unsupervised machine learning on graphs
- lammalamma25 7y agoVery cool. This kind of work makes python and ML better for everyone. I'm not sure I have a use case at the moment, but I'll definitely be keeping an eye on this one.
- cl42 7y agoThis is a great name!! For those who aren't familiar, there's a famous graph data set about a karate club: https://en.wikipedia.org/wiki/Zachary%27s_karate_club https://en.wikipedia.org/wiki/Zachary%27s_karate_club Thanks for bringing me back to undergrad... :) (...where I learned about this data set.)
- teruakohatu 7y agoThank you for sharing. The DOI is 10.2307/3629752 if anyone wants to look it up.
- lasershow 7y ago+1 on the great name and thanks for the reference link
- fudged71 7y agoI've recently been looking into community detection for NetworX graphs for my collection of notes. How does performance compare with other community detection packages for NetworkX like Python-Louvain?
- benitorosenberg 7y agoPerformance measured by what? NMI? Modularity?
- fudged71 7y agoNot super familiar with the measures, I'm just curious how it compares to other available approaches
- benitorosenberg 7y agoLouvain greedily optimizes modularity - it is hard to beat it in terms of modularity.
- kugestu 7y agoRandom side note: nowadays 'Leiden' is a recommended improvement on 'Louvain', fixing some connectivity issues and being faster: https://www.nature.com/articles/s41598-019-41695-z https://www.nature.com/articles/s41598-019-41695-z https://github.com/vtraag/leidenalg https://github.com/vtraag/leidenalg