已入深夜,您辛苦了!由于当前在线用户较少,发布求助请尽量完整地填写文献信息,科研通机器人24小时在线,伴您度过漫漫科研夜!祝你早点完成任务,早点休息,好梦!

A Survey on Learning to Reject

正确性 计算机科学 校准 过度自信效应 人工智能 过程(计算) 低信心 机器学习 匹配(统计) 心理学 社会心理学 统计 算法 数学 操作系统
作者
Xu-Yao Zhang,Guo-Sen Xie,Xiuli Li,Tao Mei,Cheng‐Lin Liu
出处
期刊:Proceedings of the IEEE [Institute of Electrical and Electronics Engineers]
卷期号:111 (2): 185-215 被引量:26
标识
DOI:10.1109/jproc.2023.3238024
摘要

Learning to reject is a special kind of self-awareness (the ability to know what you do not know), which is an essential factor for humans to become smarter. Although machine intelligence has become very accurate nowadays, it lacks such kind of self-awareness and usually acts as omniscient, resulting in overconfident errors. This article presents a comprehensive overview of this topic from three perspectives: confidence, calibration, and discrimination. Confidence is an important measurement for the reliability of model predictions. Rejection can be realized by setting thresholds on confidence. However, most models, especially modern deep neural networks, are usually overconfident. Therefore, calibration is a process to ensure confidence matching the actual likelihood of correctness, including two approaches: post-calibration and self-calibration. Calibration reflects the global characteristic of confidence, and the local distinguishing property of confidence is also important. In light of this, discrimination focuses on the performance of accepting positive samples while rejecting negative samples. As a binary classification problem, the challenge of discrimination comes from the missing and nonrepresentativeness of the negative data. Three discrimination tasks are comprehensively analyzed and discussed: failure rejection, unknown rejection, and fake rejection. By rejecting failures, the risk could be controlled especially for mission-critical applications. By rejecting unknowns, the awareness of the knowledge blind zone would be enhanced. By rejecting fakes, security and privacy could be protected. We provide a general taxonomy, organization, and discussion of the methods for solving these problems, which are studied separately in the literature. The connections between different approaches and future directions that are worth further investigation are also presented. With a discriminative and calibrated confidence, learning to reject will let the decision-making process be more practical, reliable, and secure.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
菌菌完成签到,获得积分10
3秒前
3秒前
maprang发布了新的文献求助10
4秒前
5秒前
陈凯完成签到,获得积分10
6秒前
cherry发布了新的文献求助20
9秒前
yunsww完成签到,获得积分10
10秒前
light完成签到 ,获得积分10
10秒前
小叶不吃香菜完成签到,获得积分10
11秒前
niufuking完成签到,获得积分10
12秒前
郑晓完成签到 ,获得积分10
13秒前
刘旺完成签到,获得积分10
13秒前
迷人的晓灵完成签到,获得积分10
14秒前
安东尼完成签到,获得积分10
17秒前
打打的应助被Bdk采纳,获得10
17秒前
今后的应助被wise111采纳,获得10
19秒前
19秒前
mnwkwcj完成签到,获得积分10
20秒前
lalala完成签到,获得积分10
21秒前
22秒前
22秒前
24秒前
共享精神的应助被krust采纳,获得30
25秒前
26秒前
DW的应助被啊啊啊采纳,获得10
26秒前
墨绾菩提完成签到,获得积分10
27秒前
Fantacy完成签到,获得积分10
27秒前
殷宵箫发布了新的文献求助10
28秒前
lalala发布了新的文献求助10
28秒前
墨绾菩提发布了新的文献求助10
29秒前
xushaojun发布了新的文献求助10
29秒前
long完成签到,获得积分10
29秒前
ytsl发布了新的文献求助10
30秒前
31秒前
爆米花的应助被乐乐采纳,获得10
33秒前
scwang完成签到 ,获得积分10
33秒前
1342314123完成签到,获得积分10
33秒前
科研通AI6.2的应助被zzzz采纳,获得10
34秒前
Bdk发布了新的文献求助10
36秒前
37秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
Grapes 1000
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8179640
求助须知:如何正确求助?哪些是违规求助? 9569910
关于积分的说明 21807611
捐赠科研通 7775123
什么是DOI,文献DOI怎么找? 3433072
关于科研通互助平台的介绍 2557032
邀请新用户注册赠送积分活动 2478768