2
$\begingroup$

Suppose I have some distribution of unweighted and weighted data. So, what is meant when the weighted data is having "10% maximum bias"? Does it mean that the weighted data matches the original correct distribution (like for example, MC Truth) 10% better than what unweighted data do?

$\endgroup$
2
  • $\begingroup$ can you give a link to an analysis where this terminology is used? $\endgroup$
    – anna v
    Commented May 24, 2018 at 19:36
  • $\begingroup$ @annav Sorry, I had heard this statement in some meeting where they were discussing some distribution for two differently weighted dataset and the MC Truth dataset. (Could be that I may have heard this statement incorrectly, but to what I remember, someone commented during the meeting that "one of the weighted dataset had 10% maximum bias"). All I know about biasing is in general like applying a trigger introduces bias and so on, ... So I was thinking about what he probably meant by his statement... $\endgroup$
    – kg__
    Commented May 24, 2018 at 21:25

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.