BriefGPT.xyz
Nov, 2023
不浪费任何一个注释：通过软标签改进单标签分类器
Don't Waste a Single Annotation: Improving Single-Label Classifiers Through Soft Labels
HTML
PDF
Ben Wu, Yue Li, Yida Mu, Carolina Scarton, Kalina Bontcheva...
TL;DR
本文主要研究了数据注释和训练方法在客观单标签分类任务中的局限性，并提出了使用软标签的方法来改善性能。
Abstract
In this paper, we address the limitations of the common
data annotation
and
training methods
for objective
single-label classification
tas
→