fwolling/PPGraw

heteroscedasticity issue

wanghaisheng opened this issue · 2 comments

heteroscedasticity is that the variance is not constant

recently I have read a paper about ppg signal processing,after a bunch of test ,the author say data collected from some unknown brand smart watch filtered ppg signal for different person is too heterogeneity so it is not usable.

do we need this in quality review of raw photoplethysmography (PPG) signals,
@fwolling

Yes, that makes sense - would be interesting to make a study specifically focusing on that! @wanghaisheng

after dig a little ,it seems quite simple , we just cal variance between diff person