I have an ECG signal and want to predict the nth instance of signal based on previous m samples.

Which of these (or some other) solution will be best suitable for it:

  1. Using past m samples as features and (m+1)th as the y
  2. Using Signal Modelling algorithms (like Padé approximant or Levinson-Durbon Recursion)

If using Signal Modelling algorithms, which algorithm (and order of poles) will be most suitable to ECG as it can vary quickly. Thanks!

  • $\begingroup$ Is 'ECG' referring to electrocardiographs or something else? $\endgroup$
    – IWS
    Apr 6, 2017 at 7:02
  • $\begingroup$ @IWS Yes Electrocardiographs $\endgroup$ Apr 7, 2017 at 10:36


Your Answer

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