If you have two different methods of calculating a continuous parameter (eg. heart rate), each with their own uncertainty, what would be some common methods of combining these parameters to create a better estimate?

One simple implementation I can think of would be a simple weighted average of the two based on a confidence metric (eg. $σ^2$) but I am curious as to what some alternative methods might be.


  • $\begingroup$ Check out this answer: dsp.stackexchange.com/a/16643/80 $\endgroup$
    – Peter K.
    Apr 3 '16 at 22:04
  • $\begingroup$ Thanks, that essentially describes my intuition for the simple weighted average. Are there any more advanced adaptive filtering/averaging strategies, even if they require some knowledge of the system? $\endgroup$ Apr 4 '16 at 8:59
  • $\begingroup$ I see this is deleted.... what do you mean by "robust"? Kalman filtering, if you know a little more about how your measurements were generated, can yield better estimates. $\endgroup$
    – Peter K.
    Apr 4 '16 at 12:01

Your Answer

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

Browse other questions tagged or ask your own question.