图 书 馆
利用平方损失学习：通过偏移Rademacher复杂度进行定位Learning with Square Loss: Localization through Offset Rademacher Complexity |
|
| 课程网址: | https://videolectures.net/videos/colt2015_liang_rademacher_comple... |
| 主讲教师: | Tengyuan Liang |
| 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: |  |
| 课程简介: | We consider regression with square loss and general classes of functions without the boundedness assumption. We introduce a notion of offset Rademacher complexity that provides a transparent way to study localization both in expectation and in high probability. For any (possibly non-convex) class, the excess loss of a two-step estimator is shown to be upper bounded by this offset complexity through a novel geometric inequality. In the convex case, the estimator reduces to an empirical risk minimizer. The method recovers the results of \citep{RakSriTsy15} for the bounded case while also providing guarantees without the boundedness assumption. Our high-probability statements for the unbounded case are based on the pathbreaking small-ball analysis of \cite{Mendelson14}. |
| 关 键 词: | 平方损失; 几何不等式; 小球分析 |
| 课程来源: | 视频讲座网 |
| 数据采集: | 2025-03-28：zsp |
| 最后编审: | 2025-03-28：zsp |
| 阅读次数: | 2 |