logistic regression coefficient interpretation

[Learn2019]

It looks like that all inputs have to be standardized before running logistic regression. My question is: how do I interpret the coefficients as well as odds ratio from the output.

Thanks!

[marten_kose]

Hi @Learn2019,

as the node description suggests, for interpretable statistics it is valid to perform a logistic regression without normalization. If you rather want to use it as a machine learning classifier, you’ll probably get better results with normalized data (and not well interpretable coefficients).

Does that help?

Best,
Marten

[Learn2019]

Marten,

Thanks!