0


具有数据相关分区的分层标签查询

Hierarchical label queries with data-dependent partitions
课程网址: https://videolectures.net/videos/colt2015_kpotufe_data_dependent_...  
主讲教师: Samory Kpotufe
开课单位: 信息不详。欢迎您在右侧留言补充。
开课时间: 2025-02-04
课程语种: 英语
中文简介:
给定一个联合分布$P_{X, Y}$在空间$\X$和标签集$\Y=\括号{0,1}$上,我们考虑用尽可能少的标签查询来恢复未标记样本的标签的问题。回收的标签可以传递给被动学习者,从而将过程转变为主动学习方法。我们基于数据的分层聚类分析了一系列标记程序。虽然过去已经研究了这种标记程序,但我们提供了一种新的参数化$P_{X, Y}$,它捕获了它们在一般低噪声设置下的行为,并解释了数据依赖聚类,从而为实际使用的工具提供了新的理论基础。
课程简介: Given a joint distribution $P_{X, Y}$ over a space $\X$ and a label set $\Y=\braces{0, 1}$, we consider the problem of recovering the labels of an unlabeled sample with as few label queries as possible. Recovered labels can be passed to a passive learner, thus turning the procedure into an active learning approach. We analyze a family of labeling procedures based on a hierarchical clustering of the data. While such labeling procedures have been studied in the past, we provide a new parametrization of $P_{X, Y}$ that captures their behavior in general low-noise settings, and which accounts for data-dependent clustering, thus providing new theoretical underpinning to practically used tools.
关 键 词: 标签集; 主动学习方法; 数据依赖聚类
课程来源: 视频讲座网
数据采集: 2025-03-30:zsp
最后编审: 2025-03-30:zsp
阅读次数: 7