Let say i have some data that follows gamma distribution, and i calculated the Mean and Standard deviation of the gamma distribution.
I also know that there are some outliers(Noise) in the data i want to exclude.
Now ,
Is it OK to take Mean+ 3 sigma to cover almost (99.7 %) all data ( like Normal distribution) ? My target is to exclude the data those are not inside 3 sigma considering as a noise. does it make sense?