NCA<T>
Models & Types ยท AiDotNet.Preprocessing.DimensionalityReduction
Neighborhood Components Analysis (NCA) for supervised dimensionality reduction.
Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> โ NCA<T>
For Beginners
NCA learns a space where k-NN works well:
- Points with same label are pulled closer together
- Points with different labels are pushed apart
- The learned transformation is linear (a matrix)
- Can be used for feature extraction before classification
Use cases:
- Preprocessing for k-NN classifiers
- Metric learning for distance-based methods
- Visualization with class structure preserved
- Feature extraction for classification tasks
How It Works
NCA is a supervised dimensionality reduction algorithm that learns a linear transformation to maximize the expected leave-one-out classification accuracy in the transformed space. It uses stochastic neighbor assignment for soft nearest-neighbor classification.
The algorithm:
- Define soft neighbor probabilities using softmax over distances
- Compute expected leave-one-out classification accuracy
- Optimize transformation matrix using gradient descent
- Project data using learned transformation
Constructors
| Constructor | Summary |
|---|---|
NCA(int nComponents, int maxIter, double learningRate, double tol, int? randomState, int[] columnIndices) |
Creates a new instance of NCA. |
Properties
| Property | Summary |
|---|---|
Matrix<T> TransformationMatrix { get; } |
Gets the learned transformation matrix. |
bool SupportsInverseTransform { get; } |
Gets whether this transformer supports inverse transformation. |
int NComponents { get; } |
Gets the number of components (dimensions). |
Methods
| Method | Summary |
|---|---|
Matrix<T> FitTransformSupervised(Matrix<T> data, int[] labels) |
Fits NCA using the provided data and labels, then transforms the data. |
string[] GetFeatureNamesOut(string[] inputFeatureNames) |
Gets the output feature names after transformation. |