Feature Selection in Python — Recursive Feature Elimination

Dario Radečić
Towards Data Science
7 min readSep 1, 2019

--

Finding optimal features to use for Machine learning model training can sometimes be a difficult task to accomplish. I’m not saying that the process itself is difficult, there are just so many methods to choose from. You’ve probably read a lot about methods like Principal Component Analysis, and it is by no means a bad method, but it just won’t tell you which features are most important — it will return principal components which are actually combinations of features (to explain in the most simple words possible).

--

--