H2O Binomial Scorer

Scores binomial classification predictions. Note that the scoring results may differ from the results of the KNIME Scorer node. Explanation: H2O optimizes the F1 score F1=2∗(Precision∗Recall)/(Precision+Recall) to find the probability threshold, which in turn is used to determine the class of a sample (positive vs. negative).

Options

Actual column
Column with actual (ground-truth) predictions.
Target class
The target class of the selected column.
Prediction column (target class)
Column with prediction probabilities of the positive class.
Output scores as flow variables
Activate to receive all scores as flow variables
Prefix of flow variables
The scores (i.e. accuracy, error rate, number of correct and wrong classification) can be exported as flow variables with a hard coded name. This option allows you to define a prefix for these variable identifiers so that name conflicts are resolved.

Input Ports

Icon
H2O Frame with actual class column and the two predicted probability columns.

Output Ports

Icon
The confusion matrix.
Icon
The accuracy statistics table.
Icon
The gains lift.

Views

Confusion Matrix
Displays the confusion matrix in a table view.

Workflows

Links

Developers

You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.