Open zickert opened 7 years ago
What type of statistical evaluation are we aiming at here?
We are aiming at comparing reco methods using FOM's, e.g. by computing a confidence interval for the difference between the values of a certain FOM for two methods.
Ps. There is an initial commit on this here: https://github.com/zickert/odl/tree/issue-1019__statistical_eval_of_fom
Ok, sounds great!
I have tried to complete this one. The work can be found on this branch, where I cherry-picked the commit from @zickert and continued. However, the branch is based on the reco_pipeline
branch, so I will wait with a PR until #1280 is on master.
Statistical evaluation using FOMs from #1017. Separate pull request in progress