"Performance =

Niki2019Niki2019 Member Posts: 1 Newbie
edited May 2019 in Help
I have used the optimize paramater on knn on a training and validation dataset to create a model.  When I apply this model to  new data, a predicted value for the labeled variable is created, however the performance (RMSE, adjusted r^2) is reported as "unknown" for the new data.  (I am able to see the performance of the original training model). 
Why is this happening?  How do I fix it?
I have attached all appropriate files and rmp files 
Tagged:

Answers

  • lionelderkrikorlionelderkrikor Moderator, RapidMiner Certified Analyst, Member Posts: 1,195 Unicorn
    Hi @Niki2019,

    totally logic ! 

    In your score set, you don't supply a "true value" for your label.
    So it's impossible for RapidMiner to compare the predictions to the true labels to in fine calculate a performance...

    I hope it is clear,

    Regards,

    Lionel
  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
    The new data you are applying the model to must have a value in the label field in order to calculate a performance.  It appears that MEDV (the label) is blank for the record you are trying to score.
    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
Sign In or Register to comment.