- 1、本文档共42页,可阅读全部内容。
- 2、有哪些信誉好的足球投注网站(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
Foundations of Machine Learning
Lecture 3
Mehryar Mohri
Courant Institute and Google Research
mohri@cims.nyu.edu
Learning Bounds for
Infinite Hypothesis Sets
Motivation
With an infinite hypothesis set H, the error bounds
of the previous lecture are not informative.
Is efficient learning from a finite sample possible
when H is infinite?
Our example of axis-aligned rectangles shows that
it is possible.
Can we reduce the infinite case to a finite set?
Project on finite samples?
Are there useful measures of complexity for
infinite hypothesis sets?
Mehryar Mohri - Foundations of Machine Learning page 3
This lecture
Rademacher complexity
Growth Function
VC dimension
Lower bound
Mehryar Mohri - Foundations of Machine Learning page 4
Rademacher Complexity
Definitions: let G be a family of functions mapping
from Z to [a, b] .
• Empirical Rademacher complexity of G :
m
1
R (G) = E sup σ g (z ) ,
S i i
σ g ∈G m i=1
where σ s are independent uniform random variables
i
taking values in {−1, +1} and S =(z , . . . , z ).
1 m
• Rademacher complexity of G :
Rm (G) = E [RS (G)].
S∼Dm
Mehryar Mohri - Foundations of Machine Learning page 5
Rademacher Complexity Bound
文档评论(0)