|
課程名稱 |
線上凸最佳化
Online Convex Optimization |
|
開課學期 |
112-1 |
|
授課對象 |
電機資訊學院 資訊工程學研究所 |
|
授課教師 |
李彥寰 |
|
課號 |
CSIE5062 |
|
課程識別碼 |
922 U4790 |
|
班次 |
|
|
學分 |
3.0 |
|
全/半年 |
半年 |
|
必/選修 |
選修 |
|
上課時間 |
星期一8,9,10(15:30~18:20) |
|
上課地點 |
資107 |
|
備註 |
總人數上限：20人 |
|
|
|
|
課程簡介影片 |
|
|
核心能力關聯 |
核心能力與課程規劃關聯圖 |
|
課程大綱
|
|
為確保您我的權利,請尊重智慧財產權及不得非法影印
|
|
課程概述 |
**This is a theory course. You will need to read and write mathematical proofs. There will not be any coding assignments.**
Online convex optimization (OCO) is an interdisciplinary topic, lying at the intersection of machine learning, game theory, and optimization. In machine learning, OCO is a theoretical paradigm that is different from yet has close relationship with statistical learning. In game theory, OCO naturally arises in learning-in-games studies and offers numerical algorithms computing game equilibria. In optimization, OCO provides new ideas in the design of stochastic and adaptive optimization algorithms; the most famous instance is perhaps AdaGrad.
This course aims to introduce basic OCO concepts and algorithms to the students. Tentative topics include
- Basic convex analysis.
- Online-to-batch conversion.
- Follow-the-leader-type algorithms.
- Online mirror descent.
- Solving min-max problems by learning dynamics.
- Adaptive online convex optimization.
The topics may change with respect to the latest development in OCO. |
|
課程目標 |
After taking this course, the students are expected to:
- Understand the relation between statistical and online learning paradigms,
- Be familiar with basic online convex optimization algorithms and their performance analyses, and
- Be able to read literature on online convex optimization. |
|
課程要求 |
The students are expected to be familiar with calculus, linear algebra, and probability. Knowledge in machine learning, convex optimization, and statistics can be helpful but is not necessary. |
|
預期每週課後學習時數 |
|
|
Office Hours |
|
|
指定閱讀 |
|
|
參考書目 |
Main reference:
- F. Orabona. A modern introduction to online learning. 2022. (arXiv:1912.13213)
Other references:
- S. Bubeck. Introduction to Online Optimization. 2011. (http://sbubeck.com/BubeckLectureNotes.pdf)
- S. Shalev-Shwartz. Online Learning and Online Convex Optimization. 2012. (https://www.cs.huji.ac.il/~shais/papers/OLsurvey.pdf)
- E. Hazan. Introduction to Online Convex Optimization. 2021. (arXiv:1909.05207) |
|
評量方式
(僅供參考) |
|
|