rfOOB: Out-of-bag performance estimation for random forests

In CORElearn: Classification, Regression and Feature Evaluation

Out-of-bag performance estimation for random forests

Description

The method returns internal out-of-bag performance evaluation for given random forests model.

Usage

  rfOOB(model)

Arguments

model

The model of type rf or rfNear as returned by CoreModel.

Details

The method returns random forest performance estimations obtained via its out-of-bag sets. The performance measures returned are classification accuracy, average classification margin, and correlation between trees in the forest. The classification margin is defined as the difference between probability of the correct class and probability of the most probable incorrect class. The correlation between models is estimated as the ratio between classification margin variance and variance of the forest as defined in (Breiman, 2001).

Value

The list containing three performance measures computed with out-of-bag instances is returned:

accuracy	the classification accuracy of the forest,
margin	the average margin of classification with the forest,
correlation	the correlation between trees in the forest.

Author(s)

Marko Robnik-Sikonja.

References

Leo Breiman: Random Forests. Machine Learning Journal, 2001, 45, 5-32

See Also

CORElearn, CoreModel.

Examples

# build random forests model with certain parameters
modelRF <- CoreModel(Species ~ ., iris, model="rf", 
              selectionEstimator="MDL", minNodeWeightRF=5, 
              rfNoTrees=100, maxThreads=1)
rfOOB(modelRF) 

destroyModels(modelRF) # clean up

CORElearn documentation built on Nov. 18, 2022, 5:08 p.m.