MiniBatchSparsePCA<T>

Models & Types · AiDotNet.Preprocessing.DimensionalityReduction

Mini-batch Sparse PCA using online dictionary learning.

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

For Beginners

Think of this as SparsePCA on a budget:

  • SparsePCA looks at ALL your data at once (memory intensive)
  • MiniBatchSparsePCA looks at small pieces at a time (memory efficient)
  • Results are similar, but mini-batch is faster for large datasets
  • Trade-off: Slightly less accurate but much more scalable

How It Works

MiniBatchSparsePCA is a faster, memory-efficient version of SparsePCA that processes data in mini-batches instead of using the full dataset. This makes it suitable for large datasets that don’t fit in memory.

The algorithm uses online dictionary learning with mini-batches, updating the components incrementally as it processes each batch.

Constructors

Constructor Summary
MiniBatchSparsePCA(int nComponents, double alpha, double ridge, int batchSize, int nIter, double tol, bool shuffle, int? randomState, int[] columnIndices) Creates a new instance of MiniBatchSparsePCA.

Properties

Property Summary
Matrix<T> Components { get; } Gets the sparse components (each row is a component).
Vector<T> Mean { get; } Gets the mean of each feature.
bool SupportsInverseTransform { get; } Gets whether this transformer supports inverse transformation.
double Alpha { get; } Gets the sparsity regularization parameter.
double Ridge { get; } Gets the ridge regularization parameter.
int BatchSize { get; } Gets the batch size.
int NComponents { get; } Gets the number of components.
int NSamplesSeen { get; } Gets the number of samples seen during fitting.

Methods

Method Summary
string[] GetFeatureNamesOut(string[] inputFeatureNames) Gets the output feature names after transformation.