mlr (version 2.18.0)

FeatSelResult: Result of feature selection.

Description

Container for results of feature selection. Contains the obtained features, their performance values and the optimization path which lead there. You can visualize it using analyzeFeatSelResult.

Arguments

Details

Object members:

learner (Learner)

Learner that was optimized.

control (FeatSelControl)

Control object from feature selection.

x (character)

Vector of feature names identified as optimal.

y (numeric)

Performance values for optimal x.

threshold (numeric)

Vector of finally found and used thresholds if tune.threshold was enabled in FeatSelControl, otherwise not present and hence NULL.

opt.path (ParamHelpers::OptPath)

Optimization path which lead to x.