Skip to content

A package for variable selection in regression based binary classification using marginal information values

Notifications You must be signed in to change notification settings

louis-vines/miv.select

Repository files navigation

miv.select

This package builds methods to aid with binary classification through logistic regression based methods (including GLMNet and GAMS), particularly focused on applications in credit scoring. It provides methods for supervised and unsupervised binning, single factor screening using information values and an aid for variable selection via marginal information values. There is scope to develop a fully automated algorithm for variable selection but this hasn't been added yet.

Clike here to view the vignette describing the functionality of the package.

Development

the bin folder provides command line executables for testing and documenting for those who prefer devevlopment at the command line.

Update documentation and namespace from a *nix shell:

$ bin/document

Run all tests from a unix shell:

$ bin/test

Run only a specific file:

$ bin/test <path/to/file>

NB tests require testthat version > 2.0.0 to run as expected.

About

A package for variable selection in regression based binary classification using marginal information values

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages