Scoring fails

jovajova Member, University Professor Posts: 5 University Professor
Hi everyone:

I get several models from Auto model, next I want validate these models with another dataset. The new dataset have the same structure (colums, domains and value ranges). I'm using Apply Model with the model from auto model and the new dataset. The problem is what the prediction results are only one class, no one the other class.

Do you have some tip or suggest how do the scoring process?

Thanks !

Best Answer

  • Options
    jovajova Member, University Professor Posts: 5 University Professor
    Solution Accepted

    I saw Auto model do great preprocessing work, and the final production data of each model was transformed, and the original int value was converted to real values, then when I scored the model, I made the same transformation (int to real) end the scoring process works fine!



  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,525 RM Data Scientist
    did you use the score_set process provided in the results you can save?
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
Sign In or Register to comment.