%DeepSee.extensions.clusters.CalinskiHarabasz

Class %DeepSee.extensions.clusters.CalinskiHarabasz Extends %RegisteredObject [ System = 4 ]

This class calculates Calinski-Harabasz index. Calinski-Harabasz use the Variance Ratio Criterion which is analogous to F-Statistics to estimate the number of clusters a given data naturally falls into. They minimize Within Cluster/Group Sum of Squares and maximize Between Cluster/Group Sum of Squares.

Validity indices are used in Cluster Validation and determination of the optimal number of clusters.

Properties

Model

Property Model As AbstractModel;

normalize

Property normalize As %Boolean [ InitialExpression = 1 ];

SubsetKey

Property SubsetKey As %Integer;

Methods

calculateForSample

Method calculateForSample(SampleSize As %Integer, Output sc As %Status) As %Double

calculate

Method calculate(Output sc As %Status) As %Double

traceB

Method traceB() As %Double

traceW

Method traceW() As %Double

GetSubsetCentroids

Method GetSubsetCentroids(Output zz)

traceBSubset

Method traceBSubset(zz) As %Double

traceWSubset

Method traceWSubset(zz) As %Double