0
$\begingroup$

I am working with Time-Series Data that has to be classified into two classes (Blue and Red) or at least Classify the data as one class (Red), I'm unable to come up with features that distinctly separate the data.

Please advise as to how do I need to approach this problem.

enter image description here

$\endgroup$
1
$\begingroup$

You could calculate the distance between time series using Dynamic Time Wrapping DTW and then you could cluster them using K-means or so. Here is a python implementation of the DTW or use the dtw package in R.

| improve this answer | |
$\endgroup$
  • $\begingroup$ My understanding of DTW is that it is done between 2 vectors. How would I implement it in this case? $\endgroup$ – anush95 Nov 28 '17 at 11:30
  • $\begingroup$ @anush95 you calculate the pairwise DTW and cluster after that. $\endgroup$ – salhin Nov 28 '17 at 11:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.