Sklearn Partial Least Squares Regression Learner (Labs)

Learns partial least squares regression implemented by scikit-learn library.

The model is trained with the selected numerical target column(s), and feature columns (can be numerical or nominal) from the input table. At least one numerical column and another numerical or nominal column is expected. By default, the rightmost numerical column is selected as the target column and all the remaining columns are selected as features.

If there are at least two numerical columns and two other numerical or nominal columns available, rightmost two numerical columns are selected as targets and all the remaining columns are selected as features by default. If there are only two numerical (or one numerical and one nominal) columns are available, the rightmost column is selected as the target and the other column is selected as the feature by default.

Input ports

1. Type: Table
   Input table

   Numerical and nominal columns can be selected as feature columns from this table, and the target columns must be numerical.

Output ports

1. Type: org.knime.python3.nodes.PythonBinaryBlobFileStorePortObject
   Trained model

   Trained partial least squares regression model.