Hi all!
The topic is ECG time series classification。The sample length of a piece of ECG data is 5000。 I'm having some problems with the choice between transferring raw data or extracted features。The raw data may have higher degree of interpretability but it is so long。So how is it reasonable to pass in these data and retain a high degree of interpretability?
The number of types of these data is 20. Just make a width like[5000,2048,128,20] ?
Hi all!
The topic is ECG time series classification。The sample length of a piece of ECG data is 5000。 I'm having some problems with the choice between transferring raw data or extracted features。The raw data may have higher degree of interpretability but it is so long。So how is it reasonable to pass in these data and retain a high degree of interpretability? The number of types of these data is 20. Just make a width like[5000,2048,128,20] ?