"Within Cluster Distances in ClusterCentroidEvalautor Vs ClusterDensityEvalautor"

Shubha

Hi,

Code and output of ClusterCentroidEvalautor:

<operator name="Root" class="Process" expanded="yes">
    <operator name="ExampleSetGenerator" class="ExampleSetGenerator">
        <parameter key="target_function"	value="sum classification"/>
        <parameter key="number_examples"	value="200"/>
        <parameter key="number_of_attributes"	value="2"/>
    </operator>
    <operator name="KMeans" class="KMeans">
    </operator>
    <operator name="ClusterCentroidEvaluator" class="ClusterCentroidEvaluator">
    </operator>
</operator>

PerformanceVector: avg_within_distance: -41.825 avg_within_distance_cluster_0: -38.826 avg_within_distance_cluster_1: -44.707 DaviesBouldin: -1.162

Code and Output of ClusterDensityEvalautor:

<operator name="Root" class="Process" expanded="yes">
    <operator name="ExampleSetGenerator" class="ExampleSetGenerator">
        <parameter key="target_function"	value="sum classification"/>
        <parameter key="number_examples"	value="200"/>
        <parameter key="number_of_attributes"	value="2"/>
    </operator>
    <operator name="ExampleSet2Similarity" class="ExampleSet2Similarity">
    </operator>
    <operator name="KMeans" class="KMeans">
    </operator>
    <operator name="ClusterDensityEvaluator" class="ClusterDensityEvaluator">
    </operator>
</operator>

PerformanceVector: Avg. within cluster distance: -804.184 Avg. within cluster distance for cluster 0: -760.009 Avg. within cluster distance for cluster 1: -846.627

Both the outputs have average within cluster distances, with different values. What are these values and why are they different? Thanks for the help...

Shubha

land

Hi Shuba,
the difference is, that the ClusterCentroidEvaluator calculates the distance inside a cluster by averaging the distance between each cluster example and the centroid, while the clusterDensityevaluator uses the average distances between all examples of a cluster.

Greetings,
  Sebastian