Icon

4. Clustering and Regression

KBL: Clustering and Regression
Workflow: Clustering and Regression This workflow shows a few more data analysisalgorithms:- a linear regression to predict hte number of hours/week given all other attribute values- a k-Means clustering to group together the mostsimilar data rows. Remember that k-Means, like Neural Networks,needs normalized data.The Statistics node mainly produces generalstatstical measures about one data column,including a roughly drawn histogram. assign new datato clustersgeneral statstraining setwritten by workflowPrepare Datatest setwritten by workflowPrepare Datapredict hours/weekon all remaining attrsMin-max normalizationin [0,1]3 clusterswith Euclideandistancerun predictions from model Cluster Assigner Statistics File Reader File Reader Linear RegressionLearner Normalizer k-Means RegressionPredictor Workflow: Clustering and Regression This workflow shows a few more data analysisalgorithms:- a linear regression to predict hte number of hours/week given all other attribute values- a k-Means clustering to group together the mostsimilar data rows. Remember that k-Means, like Neural Networks,needs normalized data.The Statistics node mainly produces generalstatstical measures about one data column,including a roughly drawn histogram. assign new datato clustersgeneral statstraining setwritten by workflowPrepare Datatest setwritten by workflowPrepare Datapredict hours/weekon all remaining attrsMin-max normalizationin [0,1]3 clusterswith Euclideandistancerun predictions from modelCluster Assigner Statistics File Reader File Reader Linear RegressionLearner Normalizer k-Means RegressionPredictor

Nodes

Extensions

Links