Dependent Feature

Dependent features, where predictor variables are interconnected, pose significant challenges for machine learning model explainability and performance. Current research focuses on developing methods to effectively handle these dependencies, including novel feature extraction techniques within deep learning architectures like RVFL networks augmented with fuzzy inference systems, and improved conditional independence testing using nearest-neighbor sampling. Addressing the impact of dependent features is crucial for enhancing the accuracy and interpretability of complex models across diverse applications, from improving explainable AI (XAI) methods to advancing scene-graph generation and unsupervised representation learning.