BootstrapFS<T>
Models & Types · AiDotNet.Preprocessing.FeatureSelection.Stability
Bootstrap Feature Selection for robust feature selection.
Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> → BootstrapFS<T>
Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>
For Beginners
Bootstrap is like asking the same question many times with slightly different data each time. If a feature is selected most of the time, it’s probably genuinely important. If it only gets selected sometimes, it might just be noise.
How It Works
Bootstrap Feature Selection uses bootstrap sampling (sampling with replacement) to estimate the stability of feature selection. Features that are consistently selected across many bootstrap samples are more reliable.
Constructors
| Constructor | Summary |
|---|---|
BootstrapFS(int nFeaturesToSelect, int nBootstrapSamples, double confidenceLevel, int? randomState, int[] columnIndices) |
Properties
| Property | Summary |
|---|---|
Vector<T> ConfidenceIntervals { get; } |
|
Vector<T> SelectionProbabilities { get; } |
|
bool SupportsInverseTransform { get; } |
|
double ConfidenceLevel { get; } |
|
int NBootstrapSamples { get; } |
|
int NFeaturesToSelect { get; } |
|
int[] SelectedIndices { get; } |
Methods
| Method | Summary |
|---|---|
Matrix<T> FitTransform(Matrix<T> data, Vector<T> target) |
|
bool[] GetSupportMask() |
|
string[] GetFeatureNamesOut(string[] inputFeatureNames) |
|
void Fit(Matrix<T> data, Vector<T> target) |