dalpozz/warping: Probability warping and calibration with undersampling
The package detect the condition under which undersampling is suppose to provide a better ranking. The posterior probability of a classifier is calibrated to account for the change in class prior due to undersampling.
Getting started
Browse package contents
Package details |
|
|---|---|
| Author | Andrea Dal Pozzolo, Olivier Caelen and Gianluca Bontempi |
| Maintainer | Andrea Dal Pozzolo <adalpozz@ulb.ac.be> |
| License | GPL (>= 3) |
| Version | 1.0 |
| URL | http://mlg.ulb.ac.be |
| Package repository | View on GitHub |
| Installation |
Install the latest version of this package by entering the following in R:
|
