Fuzzy Feature Selector

Optimized fuzzy-discernibility-based feature selector for large datasets using multiprocessing + numba for parallel CPU execution.

Installation

pip install fdfs

For numba acceleration (optional):

pip install fdfs[numba]

Usage

fdfs --config config.yaml

Python API

from fdfs import EnhancedOptimizedFeatureSelector

selector = EnhancedOptimizedFeatureSelector(
    df=your_dataframe,
    decision_col="target",
    use_fuzzy=True,
    n_jobs=4
)
reduced_df, selected_features = selector.select_features(best_features=10)