The DTW Wikipedia article puts this method in the category "Machine learning algorithms". On the other hand, the famous paper "Dynamic Programming Algorithm Optimization for Spoken Word Recognition" was published in IEEE Transactions on Acoustics, Speech, and Signal Processing, indicating that DTW might have been considered a signal processing algorithm by the authors.

Admittedly, the term machine learning might not have been as common in 1978 as it is today but I still wonder the following: Is DTW considered to be a machine learning algorithm or a signal processing algorithm?

  • $\begingroup$ Some teach PCA as a ML tool $\endgroup$
    – Aksakal
    Commented Apr 30, 2019 at 15:21
  • 1
    $\begingroup$ Thanks for your reply but I was asking about DTW, not PCA. $\endgroup$
    – Hagbard
    Commented Apr 30, 2019 at 15:24
  • $\begingroup$ You can anything a machine learning algorithm $\endgroup$
    – Aksakal
    Commented Apr 30, 2019 at 15:25

1 Answer 1


I don't think there is an accepted answer to this question. (Which would be ironic if you were to accept an answer in this thread.) Signal processers and machine learners will both use DTW in their work and consider it "theirs".

There are 170 hits on CrossValidated and 83 results on SignalProcessing, but of course the two sites have very different traffic, and DTW is probably more important to signal processing "as a whole" than to statistics "as a whole".


Your Answer

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

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