Lrnr_sl: SuperLearner Algorithm

Description Usage Format Value Parameters Common Parameters See Also

Description

Learner that encapsulates the Super Learner algorithm. Fits metalearner on cross-validated predictions from learners. Then forms a pipeline with the learners.

Usage

1

Format

R6Class object.

Value

Learner object with methods for training and prediction. See Lrnr_base for documentation on learners.

Parameters

learners

The "library" of learners to include

metalearner

The metalearner to be fit on predictions from the library.

folds=NULL

An origami folds object. If NULL, folds from the task are used.

keep_extra=TRUE

Stores all sub-parts of the SL computation. When set to FALSE the resultant object has a memory footprint that is significantly reduced through the discarding of intermediary data structures.

...

Not used.

Common Parameters

Individual learners have their own sets of parameters. Below is a list of shared parameters, implemented by Lrnr_base, and shared by all learners.

covariates

A character vector of covariates. The learner will use this to subset the covariates for any specified task

outcome_type

A variable_type object used to control the outcome_type used by the learner. Overrides the task outcome_type if specified

...

All other parameters should be handled by the invidual learner classes. See the documentation for the learner class you're instantiating

See Also

Other Learners: Custom_chain, Lrnr_HarmonicReg, Lrnr_arima, Lrnr_bartMachine, Lrnr_base, Lrnr_bilstm, Lrnr_condensier, Lrnr_cv, Lrnr_dbarts, Lrnr_define_interactions, Lrnr_expSmooth, Lrnr_glm_fast, Lrnr_glmnet, Lrnr_glm, Lrnr_grf, Lrnr_h2o_grid, Lrnr_hal9001, Lrnr_independent_binomial, Lrnr_lstm, Lrnr_mean, Lrnr_nnls, Lrnr_optim, Lrnr_pca, Lrnr_pkg_SuperLearner, Lrnr_randomForest, Lrnr_ranger, Lrnr_rpart, Lrnr_rugarch, Lrnr_solnp_density, Lrnr_solnp, Lrnr_stratified, Lrnr_subset_covariates, Lrnr_svm, Lrnr_tsDyn, Lrnr_xgboost, Pipeline, Stack, define_h2o_X, undocumented_learner


jeremyrcoyle/sl3 documentation built on Dec. 6, 2018, 7:15 p.m.