0 ×

04_​Looping_​for_​Multiple_​Target_​Prediction

Workflow

Multiple Target Prediction
This workflow demonstrates the prediction of multiple targets. The first loop selects one target column per loop. You need to set the target column in the learner to the flowvariable of the column name. The second loop applies the pmml to unseen data. And compines all predictions in one file.
Learning one learner per target column Predict each predictor Multiple Target Prediction This workflow demonstrates the prediction of multiple targets. The first loop selects one target column per loop. You need to set the target column in the learner to theflowvariable of the column name. The second loop applies the pmml to unseen data. And compines all predictions in one file. loop over targetsNode 10Node 11Node 12Node 14Node 16Node 17Node 18Node 19join original dataNode 21Node 22 Column ListLoop Start RProp MLP Learner PMML To Cell Loop End Chunk Loop Start Cell To PMML Generate some data Partitioning MultiLayerPerceptronPredictor ReferenceColumn Filter Joiner Loop End (ColumnAppend) Cache Learning one learner per target column Predict each predictor Multiple Target Prediction This workflow demonstrates the prediction of multiple targets. The first loop selects one target column per loop. You need to set the target column in the learner to theflowvariable of the column name. The second loop applies the pmml to unseen data. And compines all predictions in one file. loop over targetsNode 10Node 11Node 12Node 14Node 16Node 17Node 18Node 19join original dataNode 21Node 22Column ListLoop Start RProp MLP Learner PMML To Cell Loop End Chunk Loop Start Cell To PMML Generate some data Partitioning MultiLayerPerceptronPredictor ReferenceColumn Filter Joiner Loop End (ColumnAppend) Cache

Download

Get this workflow from the following link: Download

Nodes

04_​Looping_​for_​Multiple_​Target_​Prediction consists of the following 20 nodes(s):

Plugins

04_​Looping_​for_​Multiple_​Target_​Prediction contains nodes provided by the following 3 plugin(s):