importance_randomForest: Feature importance of random forest.

In emil: Evaluation of Modeling without Information Leakage

Description

Feature importance of random forest.

Usage

importance_randomForest(object, type = 1, ...)

Arguments

object	Fitted randomForest classifier
type	Importance can be assessed in two ways: 1. Permuted out-of-bag prediction error (default). This can only be used if the classifier was fitted with argument prediction=TRUE which is default. 2. Total decrease in node impurity.
...	Ignored.

Value

An prediction vector with elements corresponding to variables.

Author(s)

Christofer Bäcklin

See Also

emil, fit_randomForest, predict_randomForest, modeling_procedure

emil documentation built on Aug. 1, 2018, 1:03 a.m.