[R] package announcement: Generalized Boosted Models (gbm)

[Greg Ridgeway]

Generalized Boosted Models (gbm)

This package implements extensions to Y. Freund and R. Schapire's AdaBoost
algorithm and J. Friedman's gradient boosting machine (aka multivariate
adaptive regression trees, MART). It includes regression methods for least
squares, absolute loss, logistic, Poisson, Cox proportional hazards/partial
likelihood, and the AdaBoost exponential loss. It handles continuous,
nominal, ordinal covariates as well as those containing missing values. This
package also includes a preliminary out-of-bag estimator for the optimal
number of iterations, graphical tools for lower dimensional projections of
the fitted surface, and a few demos of example gbm sessions.

gbm 1.0 will soon appear on CRAN. Earlier versions have been up for a few
months and the latest includes many of the suggestions and fixes sent to me
by the early adopters.

Enjoy!

Greg

_______________________________________________________________
Greg Ridgeway, Ph.D.
Statistician
RAND
http://www.rand.org/methodology/stat/

_______________________________________________
R-announce at stat.math.ethz.ch mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-announce