How to recode dummy coded variables to useful data?

filanfilan Member Posts: 11 Contributor I
edited November 2018 in Help

Dear Data Mining and Rapid Miner Experts,

 

I have a dataset that contains both numerical and nominal data and I want to do linear regression with it. The thing is, I am not sure how to handle nominal data in this case.

 

I have tried to convert nominal data to numerical data but how can I recode the dummy coded variables to useful data so that I can use it in my Linear Regression operator (as in combine with the process that I have done with numerical data) to generate performance?

Answers

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn

    If you have more information on what the dummy values originally corresponded to, you can use the Map operator or Replace operator or Generate Attributes operator to repopulate the data with numerical values.  Of course, that only applies if the original nominal categories corresponded to ranges of a numeric variable, for instance.  If they are truly nominal in nature (e.g., unordered categories) then the dummy attribute coding for each category is the best form for them in a regression framework.

     

     

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