How to save an optimized predictive model and use in a NEW process with NEW data

adamfadamf Member Posts: 34 Contributor I
edited November 2018 in Help



After creating a process that uses cross-validation and optimization to train and generate an optimized predictive model, I would like to SAVE the model (to the repository?) so that I can use it in a NEW process to analyze and make predictions with NEW data.  What is the procedure to do this?


Thank you,



  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    @adamf Yes, you need to save the model in a repository and then get your scoring data.  Load in your scoring data, use an Apply Model operator, and connect in your Model. Then run it and the output will have the predictions. 

  • adamfadamf Member Posts: 34 Contributor I

    Thanks Thomas.  To save the model and then use it in another process, I would use the Store and Retrieve operators, respectively, taking the "mod" output port from the model operator (or in my case, the Cross Validation operator)?


    I'm not clear what you mean by "get your scoring data".  When I save the model, am I saving the specifically trained and optimized model based on my original process (which included nested Optimization and Cross Validation operators)?




  • BalazsBaranyBalazsBarany Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert Posts: 955 Unicorn



    "Scoring data" is the new data. That's the table that you'd like to score with your prepared model. 

    Usually the scoring data don't have labels - you get the predictions by applying the model on them.




Sign In or Register to comment.