PMML exported using One to Many (PMML) is invalid

KNIME Analytics Platform
#1

I tried to predict using PMML that exported using One to Many (PMML), but doesn't works with external predictor.(e.g. JPMML, JPMML-Spark)

I think this is because of the fields that One to Many to applies are not defined as MiningField, So it can not get the value to apply.
This behavior occurs in all Transformation Nodes (e.g. Numeric Binner).

Quoting the use of PMML 4.2:


> All data entering a model must pass through the MiningSchema. 
http://dmg.org/pmml/v4-2-1/MiningSchema.html


Also, as an implementation of JPMML, it is assumed that the field to which Transformation is applied is defined as MiningField.

Thank you.

#2

Hello Harigae,

May I ask you what you are trying to predict?

If you could provide an example that would help to understand better what you are trying to do.

The JPMML node is a general PMML predictor for classification problem. In this case, KNIME Analytics Platform rely on a specific learner-predictor motif (see attachment).Moreover, in attachment, you can find the data mining process overview that you should follow in order to make predictions.

Hope that helps,

Best,

Vincenzo