Logistic Regression - how to predict polynominal attribute type?

LeMarcLeMarc Member Posts: 72 Contributor II
edited April 2020 in Help

I would like to use Logistic Regression to predict a polynominal attribute. I know Logistic Regression is meant to be used for binominal labels. However Im still interested in exploring this algorithm with a polynominal attribute type. So therefore the selected attribute ist convert from polynominal to binominal. In order to apply the model (LR) each extra created attribut needs to be set as "label". If a polynominal attribute has only three different values thats fine. What if I want to use an Attribute with lets say more than 15 different available values. Its cumbersome to select each created attribute as a label.

My question: Is there an easier way how to use LR for predicting polynominal attributes?


Best Answers

  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,525 RM Data Scientist
    Solution Accepted

    maybe you want to have a look on the operator "Polynominal by Binomal Classification"

    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    jacobcybulskijacobcybulski Member, University Professor Posts: 391 Unicorn
    Solution Accepted
    Another possibility is to use a General Linear Model with multinomial family type. 


Sign In or Register to comment.