search

mljar / mljar-supervised

Python package for AutoML on Tabular Data with Feature Engineering, Hyper-Parameters Tuning, Explanations and Automatic Documentation
https://mljar.com
MIT License
2.98k stars 394 forks source link

Add ridge and lasso regression #446

Open fortierq opened 3 years ago

fortierq commented 3 years ago

It seems to me that there is no regularized linear regression among the available algorithms, which are useful in my experience.
I can try to add Ridge and/or Lasso, if you think it is worth it?

pplonski commented 2 years ago

@fortierq good idea, they can be added and left as a custom choice (can be selected in algorithms parameter in AutoML() constructor). Both algorithms can be added as they are from scikit-learn which is already in use.

If you are interested in adding both algorithms that would be fantastic! They should be add in similar way as Linear algorithm (code), example how to set parameters for tuning (Random Forest code).