KOBT: Knockoff Boosted Tree

A novel strategy for conducting variable selection without prior model topology knowledge using the knockoff method (Barber and Candes (2015) <doi:10.1214/15-AOS1337>) with extreme boosted tree models (Chen and Guestrin (2016) <doi:10.1145/2939672.2939785>). This method is inspired by the original knockoff method, where the differences between original and knockoff variables are used for variable selection with false discovery rate control. In addition to the original knockoff generating methods, two new sampling methods are available to be implemented, namely the sparse covariance and principal component knockoff methods. As results, the indices of selected variables are returned.

Getting started

Package details

AuthorTao Jiang [aut, cre]
MaintainerTao Jiang <tjiang8@ncsu.edu>
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:

Try the KOBT package in your browser

Any scripts or data that you put into this service are public.

KOBT documentation built on Feb. 20, 2020, 5:08 p.m.