gbt.fit: Fits a gbt model. Not intended for end-user. Shouldn't be...
In harrysouthworth/gbt: Gradient Boosted Trees

Description Usage Arguments Details Value Author(s)

Fits gradient boosted trees.

1	gbt.fit(formula, loss = "squaredLoss", data, n.trees, interaction.depth, shrinkage, bag.fraction, cv.folds, conjugate.gradient, store.results, verbose)

`formula`	a symbolic description of the model to be fit
`loss`	a character string specifying the name of the loss function to use. Currently available options are "squaredLoss" for regression, and "binomialDeviance" for binary classification.
`data`	a data frame containing the variables in the model. By default the variables are taken from `environment(formula)`, typically the environment from which `gbt` is called. Examples with missing values are ignored.
`n.trees`	the total number of trees to fit. This is equivalent to the number of iterations and the number of trees in the additive expansion.
`interaction.depth`	The maximum depth of variable interactions. 1 implies an additive model, 2 implies a model with up to 2-way interactions, etc.
`shrinkage`	a shrinkage parameter way of evolving applied to the trees in the expansion. Also known as the learning rate or step-size reduction. Currently available options are: list(type="fixed", value=v) where v is the fixed learning rate list(type="arithmetic", start=v1, end=v2) where v1 is the learning rate of the first iteration and v2 of the last. In between, the learning rate follows an arithmetic decrease (constant step size). list(type="geometric", start=v1, end=v2) where v1 is the learning rate of the first iteration and v2 of the last. In between, the learning rate follows an geometric decrease. list(type="negative.exp", start=v1, end=v2, iter75=m) where v1 is the learning rate of the first iteration and v2 of the last. iter75 is the iteration number at which the learning rate will have reached 75
`bag.fraction`	the fraction of the training set observations randomly selected to propose the next tree in the expansion. This introduces randomnesses into the model fit. If `bag.fraction`<1 then running the same model twice will result in similar but different fits. `gbt` uses the R random number generator so `set.seed` can ensure that the model can be reconstructed.
`cv.folds`	Number of cross-validation folds to perform. If `cv.folds`>1 then `gbm`, in addition to the usual fit, will perform a cross-validation.
`conjugate.gradient`	Will perform a conjugate gradient descent using the Polak-Ribiere rule instead of the classic gradient descent. This functionality is experimental.
`store.results`	If 1, the results (train (and test, in case of cross-validation) deviances, and step sizes, for each cross-validation) as displayed when `verbose` is activated are stored in `gbt$results`. They can be used to produce an estimate of the generalization error in order to select the optimal number of iterations.
`verbose`	If 1, will display results about each model that is built, once it is built.