Icon

Telecom Churn--Demonstrating kmeans++ effectiveness

Telecom Churn--Demonstrating kmeans++ effectiveness
How effective kmeans++ is--A Demo1. This example clearly shows How effective kmeans++ is in initial selection of good (random) centroids.2. Refer: https://www.geeksforgeeks.org/ml-k-means-algorithm/ for how kmeans++ works3. With kmeans++ we get very high Silhoutte Coefficients. Without using kmeans++, Silhoutte Coefficients are very low.4.See Moodle as to how Silhoutte Coefficients are calculated for each data-point: http://203.122.28.230/moodle/mod/resource/view.php?id=3813 ContactData.csvJoinonCustomerIDCallsData.xlsremove1.customerID2.Area(s)3.Phone(s)4.StateTransform1.Churn2.IntlPlan3.VMailPlanto stringmin-maxnormaliseall featuresNo of clusters: 3iterations: 500Rest: DefaultColourThree ClustersNode 101.vmail Message2.DayCharges3.IntlChargeshighest: 0.386lowest: -0.039No of clusters: 3use kmeans++ = TrueUse euclidean distno configurationchangeshighest: 0.659lowest:0.301'Winner-Cluster' toString CSV Reader Joiner Excel Reader Column Filter Number To String Normalizer k-Means Color Manager 2D/3D Scatterplot 3D ScatterPlot (Plotly) SilhouetteCoefficient SimpleKMeans (3.7) Weka ClusterAssigner (3.7) SilhouetteCoefficient Number To String How effective kmeans++ is--A Demo1. This example clearly shows How effective kmeans++ is in initial selection of good (random) centroids.2. Refer: https://www.geeksforgeeks.org/ml-k-means-algorithm/ for how kmeans++ works3. With kmeans++ we get very high Silhoutte Coefficients. Without using kmeans++, Silhoutte Coefficients are very low.4.See Moodle as to how Silhoutte Coefficients are calculated for each data-point: http://203.122.28.230/moodle/mod/resource/view.php?id=3813 ContactData.csvJoinonCustomerIDCallsData.xlsremove1.customerID2.Area(s)3.Phone(s)4.StateTransform1.Churn2.IntlPlan3.VMailPlanto stringmin-maxnormaliseall featuresNo of clusters: 3iterations: 500Rest: DefaultColourThree ClustersNode 101.vmail Message2.DayCharges3.IntlChargeshighest: 0.386lowest: -0.039No of clusters: 3use kmeans++ = TrueUse euclidean distno configurationchangeshighest: 0.659lowest:0.301'Winner-Cluster' toString CSV Reader Joiner Excel Reader Column Filter Number To String Normalizer k-Means Color Manager 2D/3D Scatterplot 3D ScatterPlot (Plotly) SilhouetteCoefficient SimpleKMeans (3.7) Weka ClusterAssigner (3.7) SilhouetteCoefficient Number To String

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: Administrator
Created at: 2021-02-28
On NodePit since: 2021-11-15
Last update: 2024-04-30
Created with KNIME version: v4.3.1
Tags: kmeanskmeans++

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!