Generalized Boosted Regression Models

Extensions to Freund and Schapire's AdaBoost algorithm, Y. Freund and R. Schapire (1997) and Friedman's gradient boosting machine, J.H. Friedman (2001) . Includes regression methods for least squares, absolute loss, t-distribution loss, quantile regression, logistic, Poisson, Cox proportional hazards partial likelihood, AdaBoost exponential loss, Huberized hinge loss, and Learning to Rank measures (LambdaMART).

gbm3: Generalized Boosted Models

Originally written by Greg Ridgeway between 1999-2003, added to by various authors, extensively updated and polished by James Hickey in 2016, survival models greatly improved by Terry Therneau in 2016, and currently maintained by Greg Ridgeway. Development is discussed at the gbm-dev Google Group.

gbm3 provides generalized boosted regression models with a newer API than the original gbm package. The package supports regression, classification, survival models, and learning-to-rank methods, with optional OpenMP parallelization in the core fitting code.

Documentation and vignettes are available at https://gbm-developers.github.io/gbm3/.

To install the development version from GitHub, first install remotes:

install.packages("remotes")

Then install gbm3:

remotes::install_github("gbm-developers/gbm3")

# or to build vignettes during installation
remotes::install_github("gbm-developers/gbm3", build_vignettes = TRUE, force = TRUE)