多标签分类：汉明损失和子集精度真的相互冲突吗？.pdf
50墨值下载
Multi-label classification: do Hamming loss and
subset accuracy really conflict with each other?
Guoqiang Wu
Department of Computer Science and Technology, Tsinghua University
G. Wu, J. Zhu. Multi-label classification: do Hamming loss and subset accuracy
really conflict with each other? NeurIPS 2020
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
1/ 11
Guoqiang Wu (Tsinghua University)
Multi-label classification
Background
Multi-Label Classification (MLC) is a fundamental task where
each instance is associated with multiple labels simultaneously.
It has plenty of applications in reality, such as text classification,
image annotation, especially the recommendation system in
E-commercial platforms.
The efficient training of big data models can benefit from the
accelerated calculation of GPU(s). We use a high-performance
server with 8 RTX 2080Ti GPUs for MLC datasets (e.g. PASCAL
VOC and NUS-WIDE).
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
2/ 11
Guoqiang Wu (Tsinghua University)
Multi-label classification
of 11
50墨值下载
【版权声明】本文为墨天轮用户原创内容，转载时必须标注文档的来源（墨天轮），文档链接，文档作者等基本信息，否则作者和墨天轮有权追究责任。如果您发现墨天轮中有涉嫌抄袭或者侵权的内容，欢迎发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。
下载排行榜
文档被以下合辑收录
评论