localModelInterpretability: Local model interpretation

In XanderHorn/autoML: Automated machine learning

Description

Fits a linear regression model to the predicted responses or probabilities of another model. Note that for classification models, the model must predict a probability else this fails.

Usage

localModelInterpretability(train, trainedModel, seed = 1991,
  sample = 0.1)

Arguments

train	[data.frame | Required] Training set on which the model was trained
trainedModel	[mlr obj | Required] MLR trained moodel object
seed	[integer | Optional] Random seed number for reproducable results. Default of 1991
sample	[numeric | Optional] A number between 0 - 1 to sub-sample the training set for faster computational time. Default of 0.1

Value

List object containing a data.frame and a plot object.

Author(s)

Xander Horn

Examples

mod <- mlr::train(makeLearner("classif.ranger", predict.type = "prob"), iris.task)
localModelInterpretability(train = iris, mod)

XanderHorn/autoML documentation built on Aug. 5, 2020, 11:45 a.m.