KernelPrincipalComponentAnalysis Constructor (IKernel, PrincipalComponentMethod, Boolean, Int32, Boolean)

public KernelPrincipalComponentAnalysis(
	IKernel kernel,
	PrincipalComponentMethod method = PrincipalComponentMethod.Center,
	bool centerInFeatureSpace = true,
	int numberOfOutputs = 0,
	bool whiten = false
)

Public Sub New ( 
	kernel As IKernel,
	Optional method As PrincipalComponentMethod = PrincipalComponentMethod.Center,
	Optional centerInFeatureSpace As Boolean = true,
	Optional numberOfOutputs As Integer = 0,
	Optional whiten As Boolean = false
)

Parameters

kernel

Type: Accord.Statistics.KernelsIKernel
The kernel to be used in the analysis.

method (Optional)

Type: Accord.Statistics.AnalysisPrincipalComponentMethod
The analysis method to perform.

centerInFeatureSpace (Optional)

Type: SystemBoolean
True to center the data in feature space, false otherwise. Default is true.

numberOfOutputs (Optional)

Type: SystemInt32
The maximum number of components that the analysis will be able to project data into.

whiten (Optional)

Type: SystemBoolean
Whether to whiten the results or not. If set to true the generatred output will be normalized to have unit standard deviation.

Reference