loading page

Delving Deep into Adversarial Attack for Multi-label Learning
  • Fengguang Su
Fengguang Su
Applied Mathematics Center of Tianjin University, Applied Mathematics Center of Tianjin University

Corresponding Author:[email protected]

Author Profile

Abstract

This research mainly focuses on adversarial attacks in multi-label learning. We analyze the shortcomings of existing research in optimization objectives and solution methods. Then, we constructed a new measure for the optimization objects based on the Jaccrd index. Furthermore, we propose a novel approach to solving the optimization problem.