AlphaInvesting<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Online

Alpha-Investing for online feature selection with FDR control.

Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> → AlphaInvesting<T>

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Think of it like managing a budget for discoveries. You start with some “discovery wealth.” Each time you test a feature, you spend some wealth. When you find a significant feature, you earn wealth back. This ensures you don’t make too many false discoveries while still finding real ones.

How It Works

Alpha-Investing is an online algorithm for controlling the false discovery rate when testing features sequentially. It maintains a “wealth” that increases when null hypotheses are rejected and decreases when tests are performed.

Constructors

Constructor Summary
AlphaInvesting(int nFeaturesToSelect, double initialWealth, double omega, double eta, int[] columnIndices)

Properties

Property Summary
Vector<T> PValues { get; }
Vector<T> TestAlphas { get; }
bool SupportsInverseTransform { get; }
double Eta { get; }
double InitialWealth { get; }
double Omega { 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)