How do I predict 100% data, not 40%?

alaaKaalaaKa Member Posts: 1 Newbie
How do I predict 100% data, not 40%? auto model
All results are required

Answers

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

    Select the model you'd like to use (the fastest, the best by accuracy etc.). Check if it has an entry "Optimal parameters". If yes, remember these.

    Then create a new process, load your data, insert a cross validation building block after the data, change the model to the one you selected, apply the parameters AutoModel found, and remember to connect the tes (Test result set) output port to the process output. This will give you all the predicted data in the cross validation.

    If you did preprocessing in AutoModel, selected input attributes or did the automatic feature engineering, you can either do that manually in your process or copy them from the process generated by AutoModel ("Open Process" => Automatic Feature Engineering). 

    Regards,
    Balázs
Sign In or Register to comment.