Feat Selectors

Different base obj choices for the FeatSelector are shown below The exact str indicator, as passed to the obj param is represented by the sub-heading (within “”) The available feat selectors are further broken down by which can work with different problem_types. Additionally, a link to the original models documentation as well as the implemented parameter distributions are shown.

binary

“rfe”

Base Class Documentation: sklearn.feature_selection.RFE

Param Distributions

0. “base rfe”

   n_features_to_select: None
   

1. “rfe num feats dist”

   n_features_to_select: Scalar(init=0.5, lower=0.1, upper=0.99).set_mutation(sigma=0.14833333333333334).set_bounds(full_range_sampling=False, lower=0.1, upper=0.99)
   

“selector”

Base Class Documentation: BPt.extensions.feat_selectors.FeatureSelector

Param Distributions

0. “random”

   mask: 'sets as random features'
   

1. “searchable”

   mask: 'sets as hyperparameters'
   

“univariate selection c”

Base Class Documentation: sklearn.feature_selection.SelectPercentile

Param Distributions

0. “base univar fs classifier”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: 50
   

1. “univar fs classifier dist”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=50, lower=1, upper=99).set_mutation(sigma=16.333333333333332).set_bounds(full_range_sampling=False, lower=1, upper=99)
   

2. “univar fs c keep more”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=75, lower=50, upper=99).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=50, upper=99)
   

3. “univar fs c keep less”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=25, lower=1, upper=50).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=1, upper=50)
   

“variance threshold”

Base Class Documentation: sklearn.feature_selection.VarianceThreshold

Param Distributions

0. “default”

   defaults only
   

regression

“rfe”

Base Class Documentation: sklearn.feature_selection.RFE

Param Distributions

0. “base rfe”

   n_features_to_select: None
   

1. “rfe num feats dist”

   n_features_to_select: Scalar(init=0.5, lower=0.1, upper=0.99).set_mutation(sigma=0.14833333333333334).set_bounds(full_range_sampling=False, lower=0.1, upper=0.99)
   

“selector”

Base Class Documentation: BPt.extensions.feat_selectors.FeatureSelector

Param Distributions

0. “random”

   mask: 'sets as random features'
   

1. “searchable”

   mask: 'sets as hyperparameters'
   

“univariate selection r”

Base Class Documentation: sklearn.feature_selection.SelectPercentile

Param Distributions

0. “base univar fs regression”

   score_func: <function f_regression at 0x7f66935c2040>
   percentile: 50
   

1. “univar fs regression dist”

   score_func: <function f_regression at 0x7f66935c2040>
   percentile: Scalar(init=50, lower=1, upper=99).set_mutation(sigma=16.333333333333332).set_bounds(full_range_sampling=False, lower=1, upper=99)
   

2. “univar fs r keep more”

   score_func: <function f_regression at 0x7f66935c2040>
   percentile: Scalar(init=75, lower=50, upper=99).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=50, upper=99)
   

3. “univar fs r keep less”

   score_func: <function f_regression at 0x7f66935c2040>
   percentile: Scalar(init=25, lower=1, upper=50).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=1, upper=50)
   

“variance threshold”

Base Class Documentation: sklearn.feature_selection.VarianceThreshold

Param Distributions

0. “default”

   defaults only
   

categorical

“rfe”

Base Class Documentation: sklearn.feature_selection.RFE

Param Distributions

0. “base rfe”

   n_features_to_select: None
   

1. “rfe num feats dist”

   n_features_to_select: Scalar(init=0.5, lower=0.1, upper=0.99).set_mutation(sigma=0.14833333333333334).set_bounds(full_range_sampling=False, lower=0.1, upper=0.99)
   

“selector”

Base Class Documentation: BPt.extensions.feat_selectors.FeatureSelector

Param Distributions

0. “random”

   mask: 'sets as random features'
   

1. “searchable”

   mask: 'sets as hyperparameters'
   

“univariate selection c”

Base Class Documentation: sklearn.feature_selection.SelectPercentile

Param Distributions

0. “base univar fs classifier”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: 50
   

1. “univar fs classifier dist”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=50, lower=1, upper=99).set_mutation(sigma=16.333333333333332).set_bounds(full_range_sampling=False, lower=1, upper=99)
   

2. “univar fs c keep more”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=75, lower=50, upper=99).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=50, upper=99)
   

3. “univar fs c keep less”

   score_func: <function f_classif at 0x7f66935c0dc0>
   percentile: Scalar(init=25, lower=1, upper=50).set_mutation(sigma=8.166666666666666).set_bounds(full_range_sampling=False, lower=1, upper=50)
   

“variance threshold”

Base Class Documentation: sklearn.feature_selection.VarianceThreshold

Param Distributions

0. “default”

   defaults only