Feature Selection
Feature selection aims to identify the most relevant subset of features from a larger dataset, improving model performance, interpretability, and efficiency. Current research emphasizes developing novel algorithms, including those based on neural networks (e.g., RelChaNet), genetic algorithms, and large language models (LLMs), to select features effectively and efficiently, often incorporating techniques like causal inference and uncertainty quantification. These advancements are crucial for various applications, such as medical diagnosis, financial prediction, and recommender systems, where reducing dimensionality and improving model explainability are paramount. The field is also actively exploring new evaluation metrics and addressing challenges like fairness and privacy in feature selection.
Papers
MACFE: A Meta-learning and Causality Based Feature Engineering Framework
Ivan Reyes-Amezcua, Daniel Flores-Araiza, Gilberto Ochoa-Ruiz, Andres Mendez-Vazquez, Eduardo Rodriguez-Tello
ControlBurn: Nonlinear Feature Selection with Sparse Tree Ensembles
Brian Liu, Miaolan Xie, Haoyue Yang, Madeleine Udell
UDRN: Unified Dimensional Reduction Neural Network for Feature Selection and Feature Projection
Zelin Zang, Yongjie Xu, Linyan Lu, Yulan Geng, Senqiao Yang, Stan Z. Li
Quantum-Enhanced Selection Operators for Evolutionary Algorithms
David Von Dollen, Sheir Yarkoni, Daniel Weimer, Florian Neukart, Thomas Bäck
Multi-Omic Data Integration and Feature Selection for Survival-based Patient Stratification via Supervised Concrete Autoencoders
Pedro Henrique da Costa Avelar, Roman Laddach, Sophia Karagiannis, Min Wu, Sophia Tsoka