Regression.Select.MinAvgErr.FitObj-class {EnsembleCV}

Class "Regression.Select.MinAvgErr.FitObj"

Description

Class containing the fit object from the "MinAvgErr" specialization of the "Regression.Select.Fit" operation in EnsembleBase package.

Objects from the Class

Objects can be created by calls of the form new("Regression.Select.MinAvgErr.FitObj", ...).

Slots

config:

Object of class "Regression.Select.Config", containing the configuration supplied to the fit operation.

est:

Object of class "ANY", containing the estimation object needed for prediction. This is a list with elements config.opt (optimal base learner configuration), error.opt (error associated with optimal configuration), and errors (vector of errors for all base learner configurations).

pred:

Object of class "RegressionSelectPred", containing the within-sample prediction, in this case the average prediction across all partitions. Note that this prediction is not used in the ecv.regression function as the ultimate training-set prediction. Instead, base learners trained on full training set (not CV style) are used for that purpose.

Extends

Class "Regression.Select.FitObj", directly.

Author(s)

Mansour T.A. Sharabiani, Alireza S. Mahani