QualityOfClustering: Measures quality of clustering at various levels

In thecomeonman/CURD: Implementation of the CURD algorithm for clustering

Description

Measures quality of clustering at various levels

Usage

QualityOfClustering(dtDatasetPredicted, cClassLabelFeat)

Arguments

dtDatasetPredicted,

a result of AssignCluster

Value

Class Node N Layer PctOfNode NodeLabel Correct LayerCumPct 1: republican Root>1 5 Layer1 0.04807692 democrat FALSE 0.021551724 2: democrat Root>2 25 Layer1 0.19531250 republican FALSE 0.129310345 3: republican Root>2 103 Layer1 0.80468750 republican TRUE 0.573275862 4: democrat Root>1 99 Layer1 0.95192308 democrat TRUE 1.000000000 5: republican Root>1>2 1 Layer2 0.02439024 democrat FALSE 0.004310345 6: democrat Root>2>2 5 Layer2 0.04629630 republican FALSE 0.025862069

thecomeonman/CURD documentation built on May 20, 2019, 7:37 a.m.