Transforms your uncalibrated Machine Learning scores to well-calibrated prediction estimates that can be interpreted as probability estimates. The implemented BBQ (Bayes Binning in Quantiles) model is taken from Naeini (2015, ISBN:0-262-51129-0).
| Version: | 0.1.1 |
| Depends: | R (≥ 2.10.0) |
| Imports: | ggplot2, pROC, reshape2, parallel, foreach, stats, fitdistrplus, doParallel |
| Published: | 2018-08-27 |
| Author: | Johanna Schwarz |
| Maintainer: | Dominik Heider <heiderd at mathematik.uni-marburg.de> |
| License: | LGPL-3 |
| NeedsCompilation: | no |
| CRAN checks: | CalibratR results |
| Reference manual: | CalibratR.pdf |
| Package source: | CalibratR_0.1.1.tar.gz |
| Windows binaries: | r-devel: CalibratR_0.1.1.zip, r-release: CalibratR_0.1.1.zip, r-oldrel: CalibratR_0.1.1.zip |
| OS X binaries: | r-release: CalibratR_0.1.1.tgz, r-oldrel: CalibratR_0.1.1.tgz |
| Old sources: | CalibratR archive |
Please use the canonical form https://CRAN.R-project.org/package=CalibratR to link to this page.