lennartwallentin / feature_selection_functions Public

Feature selection is widely used in nearly all data science pipelines. Hence I have created functions that do a form of backward stepwise selection based on the XGBoost classifier feature importance and a set of other input values with the goal to return the number of features to keep in regard to a prefered AUC-score.

0 stars 0 forks Branches Tags Activity

Notifications

Error
Looks like something went wrong!

About

Feature selection is widely used in nearly all data science pipelines. Hence I have created functions that do a form of backward stepwise selection based on the XGBoost classifier feature importance and a set of other input values with the goal to return the number of features to keep in regard to a prefered AUC-score.

python machine-learning automation pipeline feature-selection xgboost feature-engineering machine-learning-pipelines interpretability feature-importance functions-python pipelines-supervised-learning

Report repository

Languages

Jupyter Notebook 100.0%