Enhancing Fuzzy Forests with Consensus Clustering for Unbiased and Robust Feature Selection
Conférence : Communications avec actes dans un congrès international
This study presents the Fuzzy Forests algorithm, which uses consensus clustering
to improve feature selection in high-dimension data and address multicollinearity
issues. While Fuzzy Forests mitigates feature selection biases, its
effectiveness relies on the clustering method used. Our proposed consensus
clustering framework enhances robustness and reduces variability in results,
demonstrating better feature independence through extensive simulations.