Step-Wise Regression

11011400011101140001 Member Posts: 3 Contributor I
edited December 2018 in Help

Anyone knows how to perform a Step-wise Regression with RM?
Which would be the right operator? or would I need to install a specific extension?

Thanks...

Tagged:

Answers

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 916   Unicorn

    If you are looking for a classic stepwise regression, you can use either Forward Selection or Backward Elimination operators, which are both part of the feature selection operator folder, and use Linear Regression as the inner learner.  You can set the selection logic either based on absolute change in performance or associated alpha values.

     

     

    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
  • 11011400011101140001 Member Posts: 3 Contributor I

    @Telcontar120 This ise step-wise regression algortihm but I don't understand,must I use forward or backward elimination. 

    Stepwise.jpg

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 916   Unicorn

    This graphic depicts forward selection.  It also envisions adding all cross-product terms (2nd order interactions) and potentially even higher-order interactions.  I'm not sure where you are getting this guidance, but it certainly doesn't represent state-of-the-art best practices for model construction in modern data science, in my view.  I'd be very wary of overfitting if you blindly dump all these interactions terms into a regression framework without doing any kind of feature engineering or feature selection.

    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
    sgenzerSGolbert
Sign In or Register to comment.