Partial label feature selection with dynamic streaming labels

doi:10.1016/j.patcog.2025.112650

Partial label feature selection with dynamic streaming labels

2025-10-27

0

PRE

AI

DOI：10.1016/j.patcog.2025.112650

Original

Original request for help

Save

Abstract

En 中文

• This is the first attempt to address feature selection in dynamic and noisy label environments. • A framework based on max-conditional-relevance and min-conditional- redundancy is proposed. • A dynamic label disambiguation strategy is designed to figure out the problem of noise labels. • Extensive experiments show that PFSSL achieves significantly competitive performance.

Partial label feature selection with dynamic streaming labels

Abstract

Journal

Organization