makeSurrogateModel: Create surrogate model for different tasks

Description Usage Arguments Value

Description

Create surrogate model for different tasks

Usage

1
2
3
4
makeSurrogateModel(measure.name, learner.name, task.ids, lrn.par.set,
  tbl.results, tbl.hypPars, tbl.metaFeatures, tbl.runTime, tbl.resultsReference,
  surrogate.mlr.lrn, min.experiments = 100, benchmark = FALSE,
  time = FALSE)

Arguments

measure.name

Name of the measure to optimize

learner.name

Name of learner

task.ids

[numeric] ids of the tasks

lrn.par.set

learner-parameter set which should include relevant bounds for flow

tbl.results

df with getMlrRandomBotResults()

tbl.hypPars

df with getMlrRandomBotHyperpars()

tbl.metaFeatures

df with getMlrRandomBotHyperpars()

surrogate.mlr.lrn

model(s) that should be used as surrogate model

min.experiments

minimum number of experiments that should be available for a dataset, otherwise the dataset is excluded

benchmark

[logical] Should a benchmark experiment with different surrogate models be executed to evaluate the performance of the surrogate models? Default is FALSE.

Value

surrogate model


ja-thomas/OMLbots documentation built on May 18, 2019, 7:15 a.m.