Comment by somecontext

Of further potential interest: This paper cites an earlier paper by Keogh and Lin with the provocative title "Clustering of time-series subsequences is meaningless", available online at https://www.cs.ucr.edu/~eamonn/meaningless.pdf