Enum ImportanceAggregationMethod 

pub enum ImportanceAggregationMethod {
    Mean,
    WeightedMean(Vec<Float>),
    Median,
    TopK(usize),
    RankBased,
    BayesianAveraging,
    PermutationBased {
        n_repeats: usize,
    },
    SHAPBased {
        background_samples: usize,
    },
}

Feature importance aggregation methods for ensembles

Variants

Mean

Simple mean of feature importances

WeightedMean(Vec<Float>)

Weighted mean based on model performance

Median

Median aggregation (robust to outliers)

TopK(usize)

Use only top-k most important features

RankBased

Rank-based aggregation

BayesianAveraging

Bayesian model averaging of importances

PermutationBased

Permutation-based importance aggregation

Fields

n_repeats: usize

SHAPBased

SHAP-style additive feature attribution

Fields

background_samples: usize

Trait Implementations

impl Clone for ImportanceAggregationMethod

fn clone(&self) -> ImportanceAggregationMethod

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ImportanceAggregationMethod

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.