Creates a binary classification evaluator.
Area under ROC
Receiver operating characteristic (ROC), or ROC curve, is a graphical plot that illustrates the performance of a binary classifier system as its discrimination threshold is varied. For more details, view Wikipedia article.
Area under PR
Typically, Precision and Recall are inversely related, but when a balance between these two metrics needs to be achieved, the area under precision-recall curve could be used. For more details, view Wikipedia article.
Precision
Precision (also called positive predictive value) is the fraction of retrieved instances that are relevant. For more details, view Wikipedia article.
Recall
Recall (also known as sensitivity) is the fraction of relevant instances that are retrieved. For more details, view Wikipedia article.
F1 Score
F1 score (also F-score or F-measure) is a measure of a test’s accuracy. It considers both precision and recall of the test to compute the score. For more details, view Wikipedia article.
Since: Seahorse 1.0.0
This operation does not take any input.
Port | Type Qualifier | Description |
---|---|---|
0 | Evaluator | An Evaluator that can be used in an Evaluate operation. |
Name | Type | Description |
---|---|---|
binary metric |
SingleChoice |
The metric used in evaluation. Possible values: ["Area under ROC", "Area under PR", "Precision", "Recall", "F1 Score"] |
raw prediction column |
SingleColumnSelector |
Valid only if binary metric = Area under ROC or binary metric = Area under PR .
The raw prediction (confidence) column. |
prediction column |
SingleColumnSelector |
Valid only if binary metric = Precision , binary metric = Recall or binary metric = F1 Score .
The prediction column created during model scoring. |
label column |
SingleColumnSelector |
The label column for model fitting. |