rf_multi: Wrapper for RF Boruta bootstrapping.

In bootfs: Derive Robust Feature Sets for Two or Multiclass Classification Problems

Description

Wrapper for RF Boruta bootstrapping.

Usage

rf_multi(X, datX, maxRuns = 500, seed = 123, bstr = 100,
		ntree=500, localImp=TRUE, rfimportance="MeanDecreaseGini",
		fs.method="rf_boruta")

Arguments

X	List of grouping factors.
datX	Data matrix.
maxRuns	Boruta argument.
seed	Random seed.
bstr	Number of bootstrap iterations.
ntree	Number of trees for randomForest call.
localImp	LocalImportance in randomForest call.
rfimportance	Which measure to use for variable selection in the randomForest call. Either MeanDecreaseGini or MeanDecreaseAccuracy.
fs.method	Which method is used, either rf_boruta for Boruta selection or rf for default randomForest.

Details

An internal function, which is usually not called directly.

Value

List with two elements: bres holding the bootstrap fits and selprobes holding the selected features for each bootstrapping run.

Author(s)

Christian Bender

References

Todo.

See Also

Todo.

Examples

## Not run: 
#Todo.

## End(Not run)

bootfs documentation built on May 2, 2019, 5:50 p.m.