讀書寫作

10/25/2026

Dynamic Programming and Reinforcement Learning (動態規劃和強化學習)

Course objective: This course introduces dynamic decision-making under uncertainty, with an emphasis on dynamic programming and reinforcement learning. Drawing on applications in business and engineering, students will learn key theories and algorithms for solving multi-stage decision problems, both with and without explicit models of the environment. Assignments and a final project provide practical experience in problem formulation, algorithm evaluation, and Python-based implementation.

Use LLM to learn *** and its impact

Research and development

Marina Favaro and Jack Clark, When AI builds itself, Anthropic

As of May 2026, more than 80% of the code we merge into Anthropic’s codebase was authored by Claude.
In the second quarter of 2026, the typical engineer was merging 8× as much code per day as they were in 2024.
On the most open-ended tasks, Claude’s success rate reached 76% in May 2026, up 50 percentage points in six months.
In this world, the pace of progress in AI development becomes determined entirely by the availability of compute (or the speed of discovering various efficiencies in algorithmic training or inference) for AI systems. Humans play a substantially diminished role in their development, likely moving most of our effort towards oversight, validation, and verification of an expanding “virtual lab” run by AI systems. We expect that systems capable of automated AI research and development would have skills that would transfer to the rest of science, allowing them to begin to revolutionize other fields.

Dimitris Bertsimas and Georgios Margaritis, Robust and Adaptive Optimization under a Large Language Model Lens, arXiv:2501.00568.

上課內容 (Teaching)

許志華，中原大學，電機資訊學院，工業與系統工程學系

Please use the school's MS team account for online sessions.

Login Account: 學號@o365st.cycu.edu.tw, student_ID@o365st.cycu.edu.tw
Please use browser-based login instead of the application program to avoid login problems.

Please be considerate/polite, and show up only during the office hours of a course. The students in the lab need quiet time for their studies and research.

講義上有的內容，請同學要先查看或聽影片。有問題或疑義，再詢問助教。
Email: If you have further questions, please email the TA. (cc (副本) teacher: chhsu135 at Gmail) TA will handle your questions or concerns. TA will discuss with me if she/he cannot resolve the issues.

2026 秋 (Fall)

Dynamic Programming and Reinforcement Learning (動態規劃和強化學習)，博士班 (PhD)

Form for you to fill in
助教：廖庭煜 (Angus)

作業研究 (上)，工業三乙

登記，課前預習和每週進度
辦公室時間
宇翔 (jemmy) ：莊敬 302，週四 4 - 6 pm (email: jemmy5588 at Gmail)
Kiran (可嵐)：Tuesday 7 - 9 pm (email: kiransughra14 at Gmail) (Please use English.)

翻轉教室 (Flipped classroom)，中原隨筆 (My days at CYCU)
學習數學的四個層次，專題和論文的製作與報告 (tips for the final project and your thesis)，上台報告的方法與建議。
課程組員互評表 (Evaluation form for team members)

Deadline: the day for the last class meeting time
Default: If no one answers, everyone will get the same score for project participation. If only 1 person fills in the form, we will use that as the group score for project participation.

The best way to contact me (email): chhsu135 AT gmail DOT com (主旨：修課名稱和班級，學號，姓名) (Subject: Class, student ID, Name)
辦公室 (My office)：(工業系館) 莊敬大樓 (Building no. 23 for ISE) 201-3，電話 (phone) 03 265 4427。辦公室時間，四 3 - 5:50 pm (歡迎使用 Microsoft team)。導生 (分組登記)。FB，Threads。

On Call (隨時待命)

Anthony S. Fauci, On Call: A Doctor’s Journey in Public Service, Viking, 2024/06/18 (Available at CYCU library.)

Nice talks on AI, geopolitics, optimization, SCM, and more

第一次注意到沈候選人的發言是在講 1999，透過市民反映的內容和各區市議員的深度訪談，找出大家最關心的問題，這就是資料科學和系統化解決問題的方法。以下這個影片也是前所未見，透過學者專家的提問，提出對未來台北市政府治理的方向和方法；我也希望我的選區有如此優質的候選人。誠如學生說的，很多人選前一套、選後一套；也的確有不少走鐘的教授，讓我們很失望。希望當選市長後，能兌現他的承諾，例如利用選舉補助款成立台北研究院，規劃都市的未來、建立可長可久的制度。

沈伯洋「市長，你給我聽好了」學者篇 (AI 字幕的錯誤不少，影片從 12 分開始)

(Plz read it first) Business Analytics Laboratory (商業分析實驗室)

(Important) Please read this webpage first if you want to join our lab and schdule an interview.

(26/9/19) Paper published (Scalable multicollinearity recovery, more)

(26/9/8) We welcome 育輔 (Dave) to the lab!

(26/9/5) 恭喜振宇 (Jeff)，榮獲 2026 第十九屆崇越論文大賞佳作獎 (the 19th TSC Thesis Award) (photos)

研究 (Research)

Journal articles:

C.-H. Hsu and T.-Y. Liao, Scalable Multicollinearity Recovery via Mixed-integer Optimization, Journal of the Chinese Institute of Engineers, 2026 (SCIE) (new)
P.-S. Chen, H.-Y. Liu, and C.-H. Hsu, Economic Order Quantity of Deteriorating Drugs with New and Redistributed Demands in a Closed-loop Supply Chain: From the Viewpoint of a Hospital Pharmacy, International Journal of Systems Science: Operations & Logistics, 2026, Vol. 13, No. 1. (SCIE)
P.-S. Chen, J. Lyu, C.-W. Chen, I.-H. Yu, and C.-H. Hsu, The Social-economic Value of a Renewable Energy Power Generation and Energy Storage System: A Case Study, International Journal of Systems Science: Operations & Logistics, 2026, Vol. 13, No. 1. (SCIE)
H.-A. Bui, C.-H. Hsu, H.-W.V. Young, Y.-Y. Chen, and Y.-A. Liou, Advanced Semi-Supervised Learning for Remote Sensing-Based Land Cover Classification in the Mekong River Delta, Vietnam, Remote Sensing, 2026, 18, 989. https://doi.org/10.3390/rs18070989. (SCIE)

下廚 (歐貝拉，Cooking)

對我而言，煮飯是離開螢幕、讓腦袋放空和眼睛休息的好方法、也可以半蹲切菜以活動筋骨。久坐也不好，起身煮飯時，至少是站着。當然，用料也比較健康。

整個過程也和做研究是一樣的，從開始的流程設計、火候的觀察與修正、觀察結果、再持續改善；菜色和配料也要持續變化，想辦法創新。煎魚的時候，為了避免想要一直翻面的缺點，就設定鬧鐘，避免人為疏失。

煮菜最花時間的，莫過於備菜和收尾，內人在家時，這些工作大多她包了，我只負責下鍋。最重要的是，夫人不斷地鼓勵與正向回饋，令人有成就感。有一次和上課的同學聊天，同學們說：「師母提供滿滿的情緒價值」，說得真貼切！

蘑菇香煎嫩雞 (Pan-fried Chicken with Mushrooms)

機率與統計的應用

生活中到處充滿著不確定性，例如吃飯排隊的等候時間、機台生產的良率、民調統計數字的分析等等。工業系也開設許多相關的課程，例如品質管制、資料分析、智慧製造、實驗設計、機器學習、人工智慧等等，以解決工商業的問題。不確定性和雜訊，也是 (電機、軟工、機械、化工、土木、生醫、材料) 工程系統中常見的問題。

許多人修這門課的時候，很痛苦 (註 1)。除了了解其應用外，有許多的網路資源：

3Blue1Brown: Probabilities of probabilities, Central limit theorem (可設定音軌和字幕的語言)

一些常聽的 Podcast 節目和培養英文聽力的方法

一些常聽的 Podcast 節目 (*)，適合坐車、(讓眼睛) 休息、睡前、或運動時聽 (Some frequently listened-to podcasts (*) suitable for listening while traveling, giving your eyes a rest, before bed, or during exercise.)

Acquired (You could look at its transcript using iPhone.)

Subject to: Dimitris Bertsimas

References for graduate students and my laboratory members

Knowledge to master for a better foundation (and future)

Tools and general:

Adi Ignatius, “We Want to Make Ourselves Better”: The HBR Interview with Bob Sternfels, Harvard Business Review, January–February 2026.

服務和計畫 (Service and Projects)

政府計畫 (Government Project)

國科會，高維度受限迴歸模型與稀疏特徵選取：快速座標下降演算法和應用 (High-Dimensional Constrained Regression Models and Sparse Feature Selection: Fast Coordinate Descent Algorithms and Applications)，主持人 (PI)，115 年度 (2026/8 - 2027/7)
國科會，公平最佳決策樹和其應用 (Fair and Optimal Decision Tree and Its Applications)，主持人 (PI)，114 年度 (2025/8 - 2026/7)

產學合作 (Industry-Academia Cooperation)

金屬工業研究發展中心 (Metal Industries Research & Development Centre)，最佳化決策於工站排程應用之研究 (Optimal Decision-making in Work Station Scheduling) (3)，主持人 (PI)，2024

改變世界的17個方程式

Ian Stewart, In Pursuit of the Unknown: 17 Equations That Changed the World, Basic Books, 2012.

李政崇譯，改變世界的17 個方程式 (第2版)，商周出版，2025

AI Alignment

Articles:

李珊珊，「從中學就開始培養 AI 天才？」前美國奧數主教練羅博深：這讓我脊背發涼，知識分子，2026.6.23 (Prof. Po-Shen Loh) (new)

知識分子：您的意思是，在高中階段就對 AI 人才進行針對性的超前培養，太早了？
羅博深：不只是太早的問題，而是這種「純技術導向」的教育結構本身就存在致命缺陷。
如果一個孩子從高中起就進入純理科的軌道，每天只跟演算法、模型和程式碼打交道，從來沒有讀過小說，從來沒有在具體的文學語境裡理解人類複雜的現實，不懂道德約束與同理心，那麼當他掌握了最尖端的技術力量，他可能會在無意間對整個社會造成無法挽回的傷害。
一個人如果擁有龐大的科技力量，卻對它的邊界和人文影響完全不了解，他根本不應該去接觸這些核心技術。這就像核武器一樣。
所以，每當我看到一個因為過早分軌、在知識結構裡完全缺失人文的理工科學生，我都會擔憂：他真的不應該繼續做前沿科技研究了。我們已經走到了一個技術極有可能危及全人類的節點，如果一個人的教育結構是這樣一條腿長、一條腿短的狀態，他就不應該被允許繼續往這條路上走。

知識分子：您曾提到，面對來自 AI 的挑戰，孩子們需要更加「Thoughtful」，能具體解釋一下嗎？
羅博深：我習慣用英文原詞來定義它，翻譯成「深思熟慮」之類的，並不太準確。
Thoughtful 的人有兩個核心特質。第一，有共情能力，體貼周到，真心希望讓別人快樂——不只是父母和朋友，也包括陌生人，這才是真正的利他精神。第二，肯真正地去思考——他們喜歡「理解」事物，而不只是「會做」。有些學生考試能拿高分，但只是機械地照著方法做，並不理解背後的邏輯。我要找的是那種喜歡用自己的獨立思維想清楚一件事、不滿足於別人給定的標準答案的人。
對人和對問題都有洞察，只要具備這兩點，一個人在未來會成長得很好。

Medha Bankhwal, Ankit Bisht, Michael Chui, Roger Roberts, and Ashley van Heteren, AI for social good: Improving lives and protecting the planet, McKinsey, 2024.
Yoshua Bengio et al., Managing extreme AI risks amid rapid progress, Science, 20 May 2024, Vol 384, Issue 6698, pp. 842-845 (arXiv)
Jiaming Ji et al., AI Alignment: A Comprehensive Survey, arXiv:2310.19852 (website)
Jack Stilgoe, Technological risks are not the end of the world, Science, 18 Apr 2024, Vol 384, Issue 6693.