BT : (Adaptive) Boosting Tree for Poisson distributed response variable.


The author thanks Julien Trufin and Michel Denuit for their support during the project.

The idea behind (A)BT is akin to the Gradient Boosting Methods. This package is therefore inspired by the gbm3 one, originally written by Greg Ridgeway (for more details, please refer to the following url). Some of the developed codes are then pretty similar.


This package is based on the original idea proposed by M. Denuit, D. Hainaut and J. Trufin. We refer the user to the following books and articles:

Package builder and maintainer

This package has been written and is currently maintained by Gireg Willame All remarks/suggestions/improvements are warmly welcome.

Details and installation

Non-production releases (bug fixes, mostly) will be released via the GitHub release workflow. To install from GitHub, first install devtools from CRAN:


Then install the BT package from GitHub: