Icon

Backward Feature Elimination Cluster

"optimal" k according to silhoutte coefficient backward feature elimination to select the features that maximize the silhoutte coefficient Attempt 01: mix backward feature elimination +"optimal" clusters. The number of clusters is generated with the node "Random Number Generator" and each K iterates with the Backward Feature Elimination Attempt 02: mix backward feature elimination +"optimal" clusters. The number of clusters is generated manually and each K iterates with the Backward Feature Elimination Problem: for each K thesilhoutte coefficient isthe same, so it seemthat the BackwardFeature elimination isnot working in the way Iconstructed the loop asis using the samecolumns for each KFor example for K=3 thesilhoutte coef is thesame 0.7946 with thecolumn "I1" Mix backward feature elimination + "optimal" clusters AggregateOne row at a timelibrary(cluster.datasets)DatasetExtract QualityZ-Scoreon clusterAppendColumnsSelectedAppend ParameterNode 24Node 25Generate kwith a number ofrecords equalK x (number of columns in data)Sort k2233....99Number ktimes de numberof columnsthat can beremovedHighestSilhouette with k=3[I1] 1 variable 0.7946 Z-Scorelibrary(cluster.datasets)Datasetk2233....99One row at a timeon clusterExtract QualityNode 38AggregateNode 41Z-Scorelibrary(cluster.datasets)DatasetHighestSilhouettewith k=5[O2, I3, O3] 3 variables 0.7739 Node 46AggregateExtract QualityHighestSilhouettewith k=3ALL variables 0.3171 Node 50Z-Scorelibrary(cluster.datasets)Datasetk23..9One row at a timeAppend ParameterAppend ParameterHighestSilhouettewith k=3[I1] 1 variable 0.7946 AppendColumnsSelectedLoop End Table Row To VariableLoop Start (deprecated) R Source (Table) Row Filter Normalizer Backward FeatureElimination Merge Variables Variable to TableColumn (deprecated) k-Means SilhouetteCoefficient Random NumbersGenerator Sorter Metanode Sorter Normalizer R Source (Table) Table Creator Table Row To VariableLoop Start (deprecated) Backward FeatureElimination Row Filter SilhouetteCoefficient Loop End k-Means Normalizer R Source (Table) Backward FeatureElimination SilhouetteCoefficient Loop End Row Filter Sorter k-Means Normalizer R Source (Table) Table Creator Table Row To VariableLoop Start (deprecated) Variable to TableColumn (deprecated) Variable to TableColumn (deprecated) Sorter Merge Variables "optimal" k according to silhoutte coefficient backward feature elimination to select the features that maximize the silhoutte coefficient Attempt 01: mix backward feature elimination +"optimal" clusters. The number of clusters is generated with the node "Random Number Generator" and each K iterates with the Backward Feature Elimination Attempt 02: mix backward feature elimination +"optimal" clusters. The number of clusters is generated manually and each K iterates with the Backward Feature Elimination Problem: for each K thesilhoutte coefficient isthe same, so it seemthat the BackwardFeature elimination isnot working in the way Iconstructed the loop asis using the samecolumns for each KFor example for K=3 thesilhoutte coef is thesame 0.7946 with thecolumn "I1" Mix backward feature elimination + "optimal" clusters AggregateOne row at a timelibrary(cluster.datasets)DatasetExtract QualityZ-Scoreon clusterAppendColumnsSelectedAppend ParameterNode 24Node 25Generate kwith a number ofrecords equalK x (number of columns in data)Sort k2233....99Number ktimes de numberof columnsthat can beremovedHighestSilhouette with k=3[I1] 1 variable 0.7946 Z-Scorelibrary(cluster.datasets)Datasetk2233....99One row at a timeon clusterExtract QualityNode 38AggregateNode 41Z-Scorelibrary(cluster.datasets)DatasetHighestSilhouettewith k=5[O2, I3, O3] 3 variables 0.7739 Node 46AggregateExtract QualityHighestSilhouettewith k=3ALL variables 0.3171 Node 50Z-Scorelibrary(cluster.datasets)Datasetk23..9One row at a timeAppend ParameterAppend ParameterHighestSilhouettewith k=3[I1] 1 variable 0.7946 AppendColumnsSelectedLoop End Table Row To VariableLoop Start (deprecated) R Source (Table) Row Filter Normalizer Backward FeatureElimination Merge Variables Variable to TableColumn (deprecated) k-Means SilhouetteCoefficient Random NumbersGenerator Sorter Metanode Sorter Normalizer R Source (Table) Table Creator Table Row To VariableLoop Start (deprecated) Backward FeatureElimination Row Filter SilhouetteCoefficient Loop End k-Means Normalizer R Source (Table) Backward FeatureElimination SilhouetteCoefficient Loop End Row Filter Sorter k-Means Normalizer R Source (Table) Table Creator Table Row To VariableLoop Start (deprecated) Variable to TableColumn (deprecated) Variable to TableColumn (deprecated) Sorter Merge Variables

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: MAURICIO
Created at: 2021-03-28
On NodePit since: 2022-02-06
Last update: 2024-04-27
Created with KNIME version: v4.3.2

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

Try NodePit Runner!