7/10/2026
(Plz read it first) Business Analytics Laboratory (商業分析實驗室)
7/01/2026
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
5/24/2026
Nonlinear Optimization (非線性最佳化)
- Course objective: Nonlinear optimization is widely used in engineering, business, data science, and machine learning. We will introduce fundamental algorithms through examples in this course, enabling students to read research articles, formulate their own problems, and solve them efficiently using the appropriate algorithms.
3/09/2026
資料科學相關研究所和職涯準備
- 資料科學:領域知識 (智慧製造,量化投資,行銷等等),數學,資訊科技
- 如果不會寫 for loop 和 function:自學 Python (每一講有作業)
- 很棒的免費課程:CS50's Introduction to Programming with Python (Malan at Harvard), Mathematics for Machine Learning and Data Science (at Coursera)
- 發展和待遇:機器學習,AI 工程師,最佳化演算法,optimization,量化交易 (方法),智慧製造演算法,作業研究 (台達電,薪水,奇點無限 ),資料科學家
- 謝謝庭煜
- 核心:IE623G 機器學習,IC258D 資料結構 (查怡老師),CS361L 資料庫系統
- 庭煜跟我念 4 + 1,修資料結構,拿到全班第一名,系上還提供一萬元的獎金。大四也有修資料庫系統,和此文章中許多難的課程,加上努力,所以研究做得好。據我所知,只有他聽進去我的建議。
- 數學:MA305G 離散數學,MA202 高等微積分(一),MA306G 數值分析
- 資訊科技:IE440G 系統模擬 (包含數學和領域知識),IE250D 網路原理與應用,IE302G 製造聯網整合技術
- 領域知識
- 作業管理 (含智慧製造):IE238G 電腦輔助設計與製造等等
- 量化交易:FA319 財務管理(一),FA295E 金融市場,FA082G 投資學等等
- 行銷:BA245 行銷管理學,BA336 網路行銷,IU209G 消費者行為 (和 IT911D 二選一) 等等
- 核心:IE326G 人工智慧導論,IE371G 實驗設計,EL602L 計算機演算法
- 數學:MA202 高等微積分(二),MA203 微分方程(一)
- 資訊科技:MA405M 圖論演算法,IE582G 視覺系統
- 領域知識
- 作業管理 (含智慧製造):IE431G 電腦整合製造,IE527G 物聯網與大數據於智慧製造應用 等等
- 量化交易:FA320 財務管理 (二),FA011 金融科技,FA323 區塊鏈與加密貨幣,FA447 期貨與選擇權等等
- 行銷:IT911D 消費者行為等等
- 獎學金和支援
- 節省一年時間。
- 學校:
- 歷年平均成績系排名在前 30% 者,發給獎學金貳萬元整。獎學金於學生就讀碩士班一年級依學期分 2 次發给。
- 全人榮譽獎,全人標竿獎。
- 系上:
- 歷年平均成績系排名在前 20% 者,發給獎學金十萬元整。
- 參與國外國際研討會與競賽之報名費、交通費、住宿費與生活費,至多兩萬元。
- 商業分析實驗室 (Business Analytics Laboratory):論文。
- 人少,老師 (每週) 單獨討論。學生的光,老師看得見。
- 根據經驗,班排不是一個絕對的指標。重點在於想要 (投入時間) 改變自己。
- 可詢問庭煜,準時畢業且有良好的研究成果。建議從三下開始;庭煜從三下準備校外機器學習競賽、暑假實習和我做研究。
- Guides for students in Business Analytics Laboratory (實驗室學生指引):修課。
- Nice quality to have: Enjoy learning and coding (or willing to), think independently, willing to share and help
- 表現優良,老師將提供研究助理費。
- 同學的問題:我們 (工業系) 從事軟體開發和資工系的差別?
- 我:如果想要從事軟體開發的 (高薪) 工作,量化學科的基礎 (例如機統、作業研究)、流程和系統的思維 (例如生產計劃與管制、品質管制、人因工程)、和商管的知識背景 (例如經濟學、會計),是工業系的優勢。只要加上軟體系統開發需要的課程,例如資料結構、資料庫、(資管) 系統分析與設計 (開放式課程)、演算法、物件導向分析與設計、作業系統、離散數學等等,就可如虎添翼。工業系畢業學長姐,目前 (2023) 擔任純網銀的董事長和台北市資訊局副局長,就是最好的例證。
- 舉例說明工業系的優勢:線上購物定價的需求函數和最佳化,資訊系統的人機介面 (可以利用機統、實驗設計、機器學習做進階分析),熱門售票活動中的 (軟體系統) 瓶頸等等。
- 曾子軒,旺宏吳敏求:AI 時代與其學程式,懂應用數學才是業界最想要人才,遠見,2023-11-10
- 一方面感念母校栽培,一方面從個人經驗中認知到統計學,甚至是人工智慧的重要,吳敏求因此效法黑石集團 (Blackstone Group) 創辦人蘇世民 (Stephen Schwarzman) 捐贈麻省理工學院蘇世民運算學院 (MIT Schwarzman College of Computing,簡稱 MIT 運算學院) 的作法,捐贈成大設置敏求智慧運算學院。
- 數學邏輯和程式語言的邏輯是相通的。
- Brian Christian and Tom Griffiths, Algorithms to Live By: The Computer Science of Human Decisions, William Collins, 2017. 甘錫安譯,決斷的演算:預測、分析與好決定的 11 堂邏輯課,行路,2017
12/18/2025
Some books and information on machine learning and AI
- 簡禎富,工業3.5:台灣企業邁向智慧製造與數位決策的戰略,天下雜誌,2019
- Alex J. Gutman and Jordan Goldmeier, Becoming a Data Head: How to Think, Speak, and Understand Data Science, Statistics, and Machine Learning, Wiley, 2021.
6/10/2025
Foundations of Computer Vision
Antonio Torralba, Phillip Isola, and William Freeman, Foundations of Computer Vision, The MIT Press, 2024.
作者是三位 MIT 的教授。如果你想了解電腦視覺 (Computer vision) 相關研究,甚至機器學習、研究方法,大力推荐。我有空就翻一下,增加對整個領域的了解,每次都有驚喜,很神奇的書。例如
4/10/2025
Some information and papers related to the seminar
(4/9/2025) 許志華 (The original speaker needs to take care of emergency affairs.)
- Dimitris Bertsimas and Michael Lingzhi Li, Scalable holistic linear regression, Operations Research Letters, Volume 48, Issue 3, May 2020, Pages 203-208.
- We could not reproduce their results in the paper, so I asked Prof. Li for his code. He told me he finished the paper as a course project (MIT style?!😂), and the codes were stored on another obsolete hard drive.
10/16/2024
Daron Acemoglu is not having all this AI hype
Robin Wigglesworth, Daron Acemoglu is not having all this AI hype, Financial Times, May 28 2024.
Who is Daron Acemoglu? 246,645 Citations so far!
8/01/2024
AI achieves silver-medal standard solving International Mathematical Olympiad problems
AlphaProof and AlphaGeometry teams, AI achieves silver-medal standard solving International Mathematical Olympiad problems, 25 JULY 2024.
4/01/2024
1 兆電晶體 GPU 的到來
鉅亨網新聞中心,台積電董事長劉德音撰文 談1兆電晶體GPU的到來,2024-03-29
文中指出,從 1997 年擊敗西洋棋人類冠軍的「深藍」,到 2023 年爆火的 ChatGPT,再過 15 年,人工智慧已經發展到可以「合成知識」(synthesize knowledge) 的地步,可以創作詩歌、創作藝術品、診斷疾病、編寫總結報告和電腦程式碼,甚至可以設計與人類製造的積體電路相媲美的積體電路。
2/05/2024
學習數學的四個層次:(3) 在許多行業的應用
2015/12/1 初稿,持續更新中。
一般性說明
- 數學是科學之母,科學則是工業的基礎,所以大學工學院的數理化課程總學分超過 1/3。可以參考如何選填大學志願。
- 應用在不同的領域 (理工商醫農、教育),如財務工程、設計電腦、貨物產銷、工程師、使用統計學分析學習成效等等。
- 抽象的模式與思考的方式,適用於現在與未來的應用,以微分為例,物理學的距離微分是速度,經濟學中成本的微分是邊際成本,電子學的電荷微分是電流。也就是說,可以使用函數表示任何待解的問題,函數的微分便可以研究其變化和極值的情況,例如機器學習中,超參數 (hyperparameter) 的學習 。
- 基本的原則變動不大,微積分、機率和統計學、和線性代數已經有 200 年以上的歷史,可幫助未來的自我學習。許多人說學校學的東西,畢業後立即過時或沒用,我覺得很疑惑。大學只是基礎教育,必須不斷地學習新的東西,以因應產業和職務的變化;最近熱門的大數據 (big data) 和人工智慧 (artificial intelligence),其數學基礎正是這些課程
。
1/24/2024
Applications of Operations Research (作業研究) (including Optimization)
為了提高同學們的學習動機,提供以下相關的資訊,以幫助同學們找到方向。也和暑期實習和未來就業中,決策支援系統中的演算法有密切關聯。以下許多的內容屬於碩博士階段的課程,也可以增加同學們就讀研究所的動機:
- Journals:
- INFORMS Journal on Applied Analytics
- INFORMS is the leading international association for Operations Research & Analytics professionals.
- The mission of INFORMS Journal on Applied Analytics is to publish manuscripts focusing on the practice of operations research and management science and the impact this practice has on organizations throughout the world.
- Good topics to be explored for the final project
- Ramayya Krishnan and Pascal Van Hentenryck, editors, Advances in Integrating AI & O.R., INFORMS EC2021, Volume 16, April 19, 2021.
1/26/2023
Bridging physics-based and data-driven modeling for COVID-19 forecasting
Rui Wang, Danielle Robinson, Christos Faloutsos, Yuyang Wang, and Rose Yu, AutoODE: Bridging physics-based and data-driven modeling for COVID-19 forecasting, NeurIPS 2020 Workshop on Machine Learning in Public Health. (best paper award at the NeurIPS Machine Learning in Public Health Workshop)
As COVID-19 continues to spread, accurately forecasting the number of newly infected, removed and death cases has become a crucial task in public health. While mechanics compartment models are widely-used in epidemic modeling, data-driven models are emerging for disease forecasting. In this work, we investigate these two types of methods for COVID-19 forecasting. Through a comprehensive study, we find that data-driven models outperform physics-based models on the number of death cases prediction. Meanwhile, physics-based models have superior performances in predicting the number of infected and removed cases. In addition, we present an hybrid approach, AutoODE, that obtains a 57.4% reduction in mean absolute errors of the 7-day ahead COVID-19 trajectories prediction compared with the best deep learning competitor.
12/06/2022
Data Science and Machine Learning
Dirk P. Kroese, Zdravko Botev, Thomas Taimre, and Radislav Vaisman, Data Science and Machine Learning: Mathematical and Statistical Methods, Chapman & Hall, 2019. (Lecture notes and Python files)
The purpose of Data Science and Machine Learning: Mathematical and Statistical Methods is to provide an accessible, yet comprehensive textbook intended for students interested in gaining a better understanding of the mathematics and statistics that underpin the rich variety of ideas and machine learning algorithms in data science.
12/03/2022
Interpretable Machine Learning
Cynthia Rudin, Chaofan Chen, Zhi Chen, Haiyang Huang, Lesia Semenova, and Chudi Zhong, Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges. Statistics Surveys, 2022.
Interpretability in machine learning (ML) is crucial for high stakes decisions and troubleshooting. In this work, we provide fundamental principles for interpretable ML, and dispel common misunderstandings that dilute the importance of this crucial topic. We also identify 10 technical challenge areas in interpretable machine learning and provide history and background on each problem. Some of these problems are classically important, and some are recent problems that have arisen in the last few years. These problems are: (1) Optimizing sparse logical models such as decision trees; (2) Optimization of scoring systems; (3) Placing constraints into generalized additive models to encourage sparsity and better interpretability; (4) Modern case-based reasoning, including neural networks and matching for causal inference; (5) Complete supervised disentanglement of neural networks; (6) Complete or even partial unsupervised disentanglement of neural networks; (7) Dimensionality reduction for data visualization; (8) Machine learning models that can incorporate physics and other generative or causal constraints; (9) Characterization of the “Rashomon set” of good models; and (10) Interpretable reinforcement learning. This survey is suitable as a starting point for statisticians and computer scientists interested in working in interpretable machine learning.
8/16/2022
靠數據賣雞蛋
作者/張紹敏 圖片來源:卓杜信,麥當勞、美芝城都是他客戶!七年級面板主管回大武山「賣雞蛋」,靠數據幫爸爸重振家業,Cheers:快樂工作人,2022/08/13
通路
為什麼要自己賣蛋?魏毓恆解釋:「傳統蛋商只要缺蛋,每天都會開車在出貨碼頭等你,不讓你把蛋給別人;可是蛋很多的時候,他們的貨車就會『連續壞一個禮拜』,遲遲不來。」產銷的不平衡,成為大武山陷入困境一大原因,而魏毓恆不打算視而不見。
7/11/2022
他高一輟學寫詩、大學6年才畢業,他是今年的菲爾茲獎得主
李忠謙,他高一輟學寫詩、大學6年才畢業,他是今年的菲爾茲獎得主:39歲的普林斯頓數學教授許埈珥,風傳媒,2022-07-06
相較於他在少年時期厭惡抵抗數學(據稱是小學一次數學考試考差所致)、甚至高中一度輟學寫詩的迷茫過去,許埈珥的人生轉折確實比在麻省理工學院拖地擦黑板的叛逆威爾還要傳奇。...
5/30/2022
學習數學的四個層次:(2) 邏輯推理和抽象思考的能力
2015/12/1 初稿,持續更新中。
一般性說明
- D. Bok, Our Underachieving Colleges, Princeton University Press, 2007. 張善楠譯,大學教了沒?,天下文化,2008。二十一世紀八個教育目標之一『思辨能力』。(註 1)
- 程式的邏輯 (if 和 for) 數十年沒有變化,學好數學有助於邏輯能力的培養,參考計算機程式補考後的人生。
3/30/2022
Efficient method for training deep networks with unitary matrices
Bobak Kiani, Randall Balestriero, Yann Lecun, and Seth Lloyd, projUNN: efficient method for training deep networks with unitary matrices, arXiv:2203.05483v2.
In learning with recurrent or very deep feed-forward networks, employing unitary matrices in each layer can be very effective at maintaining long-range stability. However, restricting network parameters to be unitary typically comes at the cost of expensive parameterizations or increased training runtime. We propose instead an efficient method based on rank-k updates -- or their rank-k approximation -- that maintains performance at a nearly optimal training runtime. We introduce two variants of this method, named Direct (projUNN-D) and Tangent (projUNN-T) projected Unitary Neural Networks, that can parameterize full N-dimensional unitary or orthogonal matrices with a training runtime scaling as O(kN^2). Our method either projects low-rank gradients onto the closest unitary matrix (projUNN-T) or transports unitary matrices in the direction of the low-rank gradient (projUNN-D). Even in the fastest setting (k=1), projUNN is able to train a model's unitary parameters to reach comparable performances against baseline implementations. By integrating our projUNN algorithm into both recurrent and convolutional neural networks, our models can closely match or exceed benchmarked results from state-of-the-art algorithms.