Feature Transformation
Feature transformation aims to improve the performance of machine learning models by mathematically altering input features into a more effective representation space. Current research focuses on automating this process, often employing reinforcement learning, evolutionary algorithms, or generative models to discover optimal transformations, sometimes within a graph-based framework. These advancements are significant because they address limitations of manual feature engineering, improving model accuracy, generalization, and interpretability across diverse applications, including image processing, federated learning, and material science.
Papers
No Re-Train, More Gain: Upgrading Backbones with Diffusion Model for Few-Shot Segmentation
Shuai Chen, Fanman Meng, Chenhao Wu, Haoran Wei, Runtong Zhang, Qingbo Wu, Linfeng Xu, Hongliang Li
Tackling Feature-Classifier Mismatch in Federated Learning via Prompt-Driven Feature Transformation
Xinghao Wu, Jianwei Niu, Xuefeng Liu, Mingjia Shi, Guogang Zhu, Shaojie Tang