create_baselearners: Create learners

In messamat/globalIRmap: What the Package Does (One Line, Title Case)

Description

Create machine learning Learner list for random forest model.

Usage

create_baselearners(in_task)

Arguments

in_task

Task of type |link[mlr3spatiotempcv]TaskClassif, |link[mlr3spatiotempcv]TaskClassifST, or |link[mlr3]TaskRegr.

Details

Learners are methods to train and predict a model for a Task and provide meta-information about the learners, such as the hyperparameters you can set. For more info on mlr3 Learners, see https://mlr3book.mlr-org.com/learners.html.

Value

This function will return different types of learners depending on the type of task input.
If a |link[mlr3spatiotempcv]TaskClassif or |link[mlr3spatiotempcv]TaskClassifST is provided, then a list is returned with six learners, a classif.ranger (standard probability random forest learner) and classif.cforest (probability conditional inference forest learner), each in three formats for class imbalance correction: 1. without any class imbalance correction (e.g., lrn_ranger), 2. with random oversampling of minority class (e.g., 'lrn_ranger_overp), and 3. with class weighting (e.g., 'lrn_ranger_weight').

If a |link[mlr3]TaskRegr is provided, then a single learner is returned of type 'regr.ranger' with maximally selected rank statistics (maxstat) as splitting rule.

For all learners, 800 trees are grown.

messamat/globalIRmap documentation built on July 4, 2021, 10:48 a.m.