Icon

2021_​03_​09_​Denormalizing_​Predictions_​after_​MLP

Denormalizing Predictions after MLP

This workflow demonstrates how to denormalize predictions after MLP is trained. Multi Layer Perceptron learner node requires that target column when numeric is normalized and Denormalizer node can not work with columns that were not used in Normalizer node. Thus it is necessary to do a bit of manipulation in order to denormalize predictions and score results properly.

Denormalizing Predictions after MLPThis workflow demonstrates how to denormalize predicitons after MLP is trained.Forum link: https://forum.knime.com/t/denormalizing-after-mlp/30755 read datasplit training setfrom test setFilter nominal/class columbuildneural networkuse neural networkto predict numeric target columnNormalize numeric columnsSplit prediction vs other columnsDenormalize input columnsDenormalize prediction column Rename prediction columnRename prediction column backAppendScore Table Reader Partitioning Column Filter RProp MLP Learner MultiLayerPerceptronPredictor Normalizer ReferenceColumn Splitter Denormalizer Denormalizer Column Rename Column Rename Column Appender Numeric Scorer Denormalizing Predictions after MLPThis workflow demonstrates how to denormalize predicitons after MLP is trained.Forum link: https://forum.knime.com/t/denormalizing-after-mlp/30755 read datasplit training setfrom test setFilter nominal/class columbuildneural networkuse neural networkto predict numeric target columnNormalize numeric columnsSplit prediction vs other columnsDenormalize input columnsDenormalize prediction column Rename prediction columnRename prediction column backAppendScore Table Reader Partitioning Column Filter RProp MLP Learner MultiLayerPerceptronPredictor Normalizer ReferenceColumn Splitter Denormalizer Denormalizer Column Rename Column Rename Column Appender Numeric Scorer

Nodes

Extensions

Links