Adversarial Multi-task Learning for Text Classification

创新点: Adversial learning + Orthogonality Constraints

Abstract

  • 背景:

    神经网络模型已经表明了它们在Multi-task Learning中的广阔前景,其重点是学习共享层以提取公共的和随不同任务不变的特征

+
Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now

×