makeStrategy: Specify your own learning strategy.

Description Usage Arguments Value See Also Examples

View source: R/Strategy.R

Description

Create a strategy which uses data about the task to automatically configure a wrapped Learner to train the learner.

Usage

1
2
makeStrategy(id, name = id, task.type = NULL, properties = NULL,
  base.learner = NULL, wrappers = NULL, fun)

Arguments

id

(character(1))
Name of the strategy.

name

(character(1))
Long name of the strategy. Default is id.

task.type

(character)
Supported task types of the strategy, e.g. 'classif', 'regr', etc.

properties

(character)
Supported learner properties of the strategy, e.g. 'numerics', 'factors', 'missings', etc.

base.learner

(character)
Base learner or learners that strategy uses

wrappers

(character)
Wrapper or wrappers that strategy uses

fun

(function(task, pars))
Takes a Task and returns a wrapped Learner customized to the task. pars may be added optionally to override default strategy settings.

task (Task)

The task around which to build the learning strategy.

pars (list)

An optional list of parameters to override the defaults.

Value

(Strategy).

See Also

strategies

Examples

1
# TBD

Prometheus77/ezmlr documentation built on May 17, 2019, 6:34 p.m.