Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Parameter-free, Dynamic, and Strongly-Adaptive Online Learning #23

Open
nocotan opened this issue Jan 15, 2021 · 0 comments
Open

Parameter-free, Dynamic, and Strongly-Adaptive Online Learning #23

nocotan opened this issue Jan 15, 2021 · 0 comments
Assignees

Comments

@nocotan
Copy link
Member

nocotan commented Jan 15, 2021

一言でいうと

最適化における時間的な区間を互いに排反な代表的区間で分割し,それらの区間におけるリグレットを抑えるようなParameter-freeのオンライン学習手法を提案.

論文リンク

http://proceedings.mlr.press/v119/cutkosky20a/cutkosky20a.pdf

著者/所属機関

Ashok Cutkosky (Google Research)

投稿日付(yyyy/MM/dd)

2020/07/12

概要

適応性のいくつかの望ましい条件を組み合わせた新しいアルゴリズムを提案.

  • 達成できるリグレット:

Screen Shot 2021-01-15 at 23 28 49

  • 強適応的リグレット:

Screen Shot 2021-01-15 at 23 30 15

  • 最適動的リグレット:

Screen Shot 2021-01-15 at 23 30 53

新規性・差分

適応性の概念を組み合わせて,parameter-freeなオンライン学習を達成する初めての手法.

手法

Screen Shot 2021-01-15 at 23 20 20

Screen Shot 2021-01-15 at 23 20 26

結果

コメント

@nocotan nocotan self-assigned this Jan 15, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant