Skip to content

suyong-choi/PositiveResampler

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 

Repository files navigation

Positive Resampler using DNN

Author: Suyong Choi (Korea University)

Positive weight resampler using DNN for samples with negative sample weights. This can be useful if you are trying to use samples that have negative weights but need them to have positive values. Such cases arise in machine learning.

In general, it is not possible to assign positive values to all the events. However, in high-energy physics, simulated events produced at next-to-leading order precision have negative weight events that have positive weight events nearby such that local average of the weight can be positive. Here, we assume that the local average weight is positive for the sample. In https://arxiv.org/abs/2005.09375, a positive resampler for binned data is introduced. However, for unbinned data, it cannot be used. Here, we turn the problem is turned into finding the weight as a function of the phase space variables.

The positivity of sample weights is imposed by using a softplus activation function at the output.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published