BaseLearner.FitObj-class {EnsembleBase}

Classes "BaseLearner.FitObj" and "Regression.FitObj"

Description

Base class templates for containing base learner training output.

Objects from the Class

"BaseLearner.FitObj" is a virtual class: No objects may be created from it. "Regression.FitObj" is a base class for objects representing trained models for individual base learners.

Slots

config:

Object of class "BaseLearner.Config"; often one of the derived configuration classes belonging to a particular base learner.

For Regression.FitObj, we have the following additional fields:

est:

Object of class "RegressionEstObj", typically containing the low-level list coming out of the training algorithm. If filemethod=TRUE during the fit, this object will be of class "character", containing the filepath to where the estimation object is stored.

pred:

Object of class "OptionalNumeric", fitted values of the model for the training data. It is allowed to be "NULL" in order to reduce memory footrpint during cross-validated ensemble methods.

Methods

No methods defined with these classes in their signature.

Author(s)

Alireza S. Mahani, Mansour T.A. Sharabiani

See Also

"KNN.Regression.FitObj", "RF.Regression.FitObj", "SVM.Regression.FitObj", "GBM.Regression.FitObj", "NNET.Regression.FitObj"