Performance of the Optimize Selection

Ledave97

Hello,
Im currently working on my first Rapid Miner Project, so I'm a total newby.
I'm try to get the best performance throug a prediction with a Neural Net,
I try to use the Optimize Selection to look for the best Attribute Combination. 
The Performance, which is given by the Optimize Selection is better then the Performance, of the applied, Model, which uses the Example Set from the Optimize Selection with the exact same Parameters in the Optimize Selection, as in the applied Model. How can I achieve this performance for my applied model? Or could I just use the performance of the Optimize Selection, as a value for the performance of the model?

I hope some can help me.

Kind regard,
David

sara20

Ledave97 ,

Hello

I recommend you to use cross validation because if you have unbalance data with cross validation you can train your machine better than split data. Also in community there is a great post from @sgenzer which can help you about cross validation.

I hope this helps
Sara

Ledave97

@sara20 Can you post a link of the post from @sgenzer, i couldn't find it.

sara20

@Ledave97
This is the link
https://community.rapidminer.com/discussion/54621/cross-validation-and-its-outputs-in-rm-studio

Good luck
Sara