summary.gbm.cverr: Summarizing gbm.cverr Results

In patr1ckm/mvtboost: Tree Boosting for Multivariate Outcomes

Description

summary method for class gbm.cverr

Usage

## S3 method for class 'gbm.cverr'
summary(object, ...)

Arguments

object	Output from gbm.cverr
...	Further arguments passed to or from other methods.

Value

Calling summary(gbm.cverr) produces a data frame with rows corresponding to sets of metaparameters and columns that denote for each row,

min.cv.error	Minimum cross-validation error resulting from the given set of metaparameters.
w.index	The index of the (optional) list of weight vectors corresponding to the given set of metaparameters. This will be omitted if a list of weights was not provided to gbm.cverr through the input parameter w.
var.monotone.index	The index of the (optional) list of monotonicity vectors corresponding to the given set of metaparameters. This will be omitted if a list of weights was not provided to gbm.cverr through the input parameter var.monotone.
interaction.depth	The interaction depth corresponding to the given set of metaparameters.
n.minobsinnode	Minimum number of observations in the terminal nodes of the trees for the given set of metaparameters.
shrinkage	The shrinkage parameter corresponding to the given set of metaparameters.
bag.fraction	The fraction of independent training observations randomly selected to propose the next tree corresponding to the given set of metaparameters.
n.trees	The optimum number of trees to utilize given the set of metaprameters denoted in the row. Note that entries in this column will be marked with '>=' if the boosting procedure was terminated due to time running out for this set of metaparameters, determined by the user-specified max.time passed to gbm.cverr

Sets of metaparameters (rows) are ordered from best (top row) to worst (last row) in terms of the resulting cross-validation error.

Author(s)

Daniel B. McArtor ([e-mail])

patr1ckm/mvtboost documentation built on May 24, 2019, 8:21 p.m.