taking off from this topic here (Hierarchical clustering, linkage methods and dynamic time warping) and from this (Dynamic Time Warping Clustering) I was wondering if anyone could tell me what could make realiable the results of running a dtw / hierarchical clustering with time series (or a sound procedure in this case). I have read that its statistical soundness might be in jeopardy so, before I stress my CPU to death for no reason, I was wondering if anyone could shed any light on the matter.

  • $\begingroup$ What did you hear was wrong with its statistical soundness? $\endgroup$ May 2, 2016 at 13:32
  • $\begingroup$ unless I got the methods mixed up (with k-means method) - which is entirely possible -, the method could have problems when series are very similar but shifted $\endgroup$
    – Asher11
    May 2, 2016 at 13:55


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.