MachineLearningkMeansClust Method

public static RetArray<long> kMeansClust(
	InArray<double> X,
	BaseArray k,
	int maxIterations = 10000,
	bool centerInitRandom = true,
	OutArray<double> outCenters = null
)

Public Shared Function kMeansClust ( 
	X As InArray(Of Double),
	k As BaseArray,
	Optional maxIterations As Integer = 10000,
	Optional centerInitRandom As Boolean = true,
	Optional outCenters As OutArray(Of Double) = Nothing
) As RetArray(Of Long)

Dim X As InArray(Of Double)
Dim k As BaseArray
Dim maxIterations As Integer
Dim centerInitRandom As Boolean
Dim outCenters As OutArray(Of Double)
Dim returnValue As RetArray(Of Long)

returnValue = MachineLearning.kMeansClust(X, 
	k, maxIterations, centerInitRandom, 
	outCenters)

Parameters

X

Type: ILNumericsInArrayDouble
Data matrix, data points are given as columns.

k

Type: ILNumericsBaseArray
Initial number of clusters.

maxIterations (Optional)

Type: SystemInt32
[Optional] Maximum number of iterations, the computation will exit after that many iterations, default: 10.000.

centerInitRandom (Optional)

Type: SystemBoolean
[Optional] false: pick the first k data points as initial centers, true: pick random datapoints (default).

outCenters (Optional)

Type: ILNumericsOutArrayDouble
[Input/Output/Optional] If not null on entry, outCenters will contain the centers of the clusters found, default: null.

Return Value

If outCenters is given not null on input, the algorithm returns the computed centers in that parameter. A matrix may be given on input, in order to give a hint of the initial center positions. This may help to find correct cluster centers - even if the initial hint is not exact. In order to do so, the matrix given must be of the correct size (X.D[0] by k) and centerInitRandom must be set to false.

[ILNumerics Machine Learning Toolbox]

Reference