Skip to content

Click in-app to access the full platform documentation for your version of DataRobot.

Normalizing for monotonicity

Robot 1

To normalize or not to normalize, that is the question

Got some questions about monotonicity (I've been testing models heavily with such constraints). Would appreciate any answers or documentation that can help.

  1. When we apply monotonic Increasing/Decreasing constraints to attributes, is DataRobot doing some kind of normalization (capping amd flooring, binning etc.)?
  2. When we apply try only monotonic models, will it try GBM, XGBOOST, RF etc.?

Thanks for the help!

Robot 2

No normalization that I know of and just xgboost and gam. You don't really need to normalize data for xgboost models.

Robot 3

For docs you can check out how to configure feature constraints and then the workflow to build them. And here's an ELI5 answer to "what does monotonic mean?"


Updated February 20, 2023
Back to top