mlr-org / mlr3proba

Probabilistic Learning for mlr3

https://mlr3proba.mlr-org.com/

GNU Lesser General Public License v3.0

130 stars 20 forks source link

Update measure docs #372

Open bblodfon opened 8 months ago

bblodfon commented 8 months ago

TODOs

surv.dcalib => p > 0.05 well d-calibrated, arbitrary, choose truncate value => Using chisq with 9 df, a model is only not D-calibrated if D>=17.
Add citation for properness RCLL, RISBS, RISLL and RISS (scoring rules paper).
Alt. name?: Calibration slope => https://mlr3proba.mlr-org.com/reference/mlr_measures_surv.calib_beta.html

Updates

407 fixed some of the above