Regression with Random Forest ?

phivuphivu MemberPosts:34Guru
edited November 2018 inHelp

嗨RapidMiner,

I'm doing regression with 480 input features. I tried to use Deep Learning operator but the training Root Mean Square Error is still quite high. Now I'm trying to use Random Forest because of its Random Subspace approach, but found that the Random Forest operator cannot handle numerical label. How can I deal with this?

Thank you very much for your support.

Best Regards,

phivu

Tagged:

Best Answers

  • earmijoearmijo MemberPosts:270Unicorn
    Solution Accepted

    你不能用RapidMiner除非你是会的ing to use R Scripts. However, the latest version of RM has a new operator Gradient Boosted Trees which is competitive with Random Forest and it can handle both numerical and polynominal labels. Explore it.

    Thomas_Ott phivu
  • earmijoearmijo MemberPosts:270Unicorn
    Solution Accepted

    Install the R Script Extension. Verify you have R installed in your computer and run the code below. I adapted the code that comes with the application to run Random Forest for a regression problem.





    <宏/ >





    Fetch example data






    Split the data in a training and a test set



    Train a RandomForest model in R and return it as an R object



    Apply the trained model on the test data












    IngoRM phivu

Answers

Sign InorRegisterto comment.