Error message in a process using Decision tree model and Cross validation operator

tonyboy9

This is my data set. The target attribute is "not.fully.paid."



This is the process page one:



This is the Process page two, Cross validation.

The error message reads: Non-nominal label. The label attribute (not.fully.paid)
must be nominal for the calculation of performance criteria for binomial classifi-
cation tools.

The Performance operator I chose is for Classification.




Any helpful advice about where I went wrong here is appreciated.

Tony

FaraazSarfalak

Hello, Mate.I think you will have to convert notfullypaid by using the operator numerical to binominal because I think the attribute notfullypaid is of integer type.

tonyboy9

Yes, thank you Faraaz. I have decided to go in another direction building the process, to rely on the CreditRiskModeling template. I have tried to include your 'Numerical to binomial' operator. It is the 'Optimize Parameters (Grid)' which concerns me.



Specifically it is the Edit Parameter Settings. I have read the Description at https://docs.rapidminer.com/latest/studio/operators/modeling/optimization/parameters/optimize_parameters_grid.html#:~:text=The Optimize Parameters (Grid) Operator,through the parameter set port.

I tried to input something here and nothing seems to work. I want to learn what options I have here for a better process. Thanks again.

Tony