This is a blog containing data related news and information that I find interesting or relevant.
Links are given to original sites containing source information for which I can take no responsibility. Any opinion expressed is my own.
Wednesday, 11 February 2015
Data Science 102: K-means clustering is not a free lunch
K-means is a widely used method in cluster analysis, but what are its underlying assumptions and drawbacks? David Robinson examines what happens for non-spherical data and unevenly sized clusters.
No comments:
Post a Comment
Note: only a member of this blog may post a comment.