Multinomial logistic regression with categorical predictors

4 views (last 30 days)
Hello,
I have (I hope) a relatively straight-forward question. I would like to use a multinomial logistic regression to predict a categorical outcome. However, I also have categorical predictors. The documentation for mnrfit doesn't seem to address categorical inputs. Is there a way to do this?
Thanks.
  1 Comment
Nirosha Rathnayake
Nirosha Rathnayake on 9 Feb 2017
Hello David, have you figure out this problem. I am also trying to work on this. Multinomial logistic regression with categorical predictors when the predictor has more than 2 categories. might have to use dummy variables (k-1) if a predictor has k categories. Which mean we should have k-1 coefficients (k-1 beta values) in the model, taking one category as a reference category. Please let me know if you have any other comments or ideas on using categorical predictors in multinomial logistic regression. Thanks.

Sign in to comment.

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!