Open continual sampling with hypersphere knowledge transfer for rapid feature selection

Xuemei Cao, Xiangkun Wang, Haoyang Liang, Bingjun Wei, Xin Yang

Abstract: Highlights•We are the first to propose open continual sampling in open environments.•We define knowledge and builds a knowledge base to manage open risks.•We propose OCS-FS, a rapid feature selection method with basic and open learning.•OCS-FS accumulates past knowledge and learns from new data without historical access.•Our method outperforms state-of-the-art feature selection on public datasets.

External IDs:dblp:journals/asc/CaoWLWY25