The following node is available in the Open Source KNIME predictive analytics and data mining platform version 2.7.1. Discover over 1000 other nodes, as well as enterprise functionality at http://knime.com.
This is the end node of a backward feature elimination loop. It compares the target column with the
prediction column set in the node's dialog. For numerical columns the
squared error is computed, for all
other columns the error rate. The column that caused the smallest error
when left out is then discarded for
the next iteration.
The node's output table shows the error for the different levels of the elimination process, and the column
that has been excluded in the level.
0 | The result table of a predictor inside the loop |
0 | Elimination levels and corresponding error rates and left out columns |
1 | The elimination levels as model that can be used in a backward elimination filter node |