.. _`Decision Function`:

.. _`org.sysess.sympathy.machinelearning.decision_function`:

Decision Function
`````````````````

.. image:: decision_function.svg
   :width: 48


Applies the decision function (if available) of a trained model to return a scalar for each class of outputs




Documentation
:::::::::::::

Predict confidence scores for samples with a model. Uses the models implementation of the
.decision_function. The decision function is typically used in regression, to get the raw
scores or distances from the decision boundary for each sample, rather than the predicted
class labels. In multi-class classification, it returns either one score per class or a score
matrix, depending on the model and configuration.


Definition
::::::::::


Input ports
...........

    **in-model**
        | Type: model
        | Description: Input model
    **X**
        | Type: table
        | Description: X

Output ports
............

    **Y**
        | Type: table
        | Description: Y















Implementation
..............

.. automodule:: node_application
    :noindex:

.. class:: DecisionFunction
    :noindex: