3 ms·
scikit-learn's materials (amazing in general) has a nice overview of their clustering methods and how they perform with some test data sets: http://scikit-learn
by a_bonobo 10y ago
scikit-learn's materials (amazing in general) has a nice overview of their clustering methods and how they perform with some test data sets: http://scikit-learn.org/stable/modules/clustering.html http://scikit-learn.org/stable/modules/clustering.html
How would you even cluster a power law distribution? There's nothing that distinguishes in two or more groups?
- hnhg 10y agoA lot of online engagement data looks like a power law distribution on large platforms. The mean (obviously) has huge variance. Say you have multiple metrics per user like this. You can still apply k-means but the huge variance makes it rather ineffective. That's what I've found, anyway.