隨機時序決策與分析
Sequential Decision Modeling and Analytics
| 節 | 週五 |
|---|---|
2 09:00–09:50 | 隨機時序決策與分析 MB506（光復） 3 節連堂 |
3 10:10–11:00 | |
4 11:10–12:00 |
* 根據陽明交大上課時間表所列
This course provides an in-depth exploration of sequential decision making (SDM) and its industrial applications. Students will learn to model SDM problems in a canonical mathematical form, apply fundamental algorithms such as dynamic programming and reinforcement learning in Python, and implement the framework in real-world applications. With homework assignments, in-class coding exercises, and a term project, students will gain both theoretical understanding and practical skills to address optimization, automation, and decision-making challenges under uncertainty.
Basic probability theory, Operations Research
無備註
Python-language programming
• Homework Assignments: 30% • Midterm Exam: 30% • Project: 40% • Participation: 5%
Markov Decision Process (MDP)
1. Basic Concept of Markov Decision Process 2. Finite or infinite MDP 3. Policy and Value Iteration Method
- 講授：
- 15
Reinforcement Learning (RL)
1. Basic concept of RL 2. Model-Free RL 3. Deep RL 4. Multi-Armed Bandits
- 講授：
- 15
Preliminary
1. Probability Theory 2. Discrete Markov Chain 3. Sequential Decision-Making Framework
- 講授：
- 9
| 週次 | 主題 |
|---|---|
| 第 1 週 | Introduction to Sequential Decision Making and Analytics 2026-09-11(五) 時數：[2026-09-11]田凱文(3.00) |
| 第 2 週 | Preliminary - Basic Probability, Conditional Probabilities 2026-09-18(五) 時數：[2026-09-18]田凱文(3.00) |
| 第 3 週 | Preliminary - Markov Chain Properties 2026-09-25(五) 時數：[2026-09-25]田凱文(3.00) |
| 第 4 週 | MDP - Sequential Decision Modeling 2026-10-02(五) 時數：[2026-10-02]田凱文(3.00) |
| 第 5 週 | MDP - Final Horizon MDP 2026-10-09(五) 時數：[2026-10-09]田凱文(3.00) |
| 第 6 週 | Holiday: Double 10th Day 2026-10-16(五) 時數：[2026-10-16]田凱文(3.00) |
| 第 7 週 | MDP - Infinite Horizon MDP 2026-10-23(五) 時數：[2026-10-23]田凱文(3.00) |
| 第 8 週 | Holiday 2026-10-30(五) 時數：[2026-10-30]田凱文(3.00) |
| 第 9 週 | Midterm Exam 2026-11-06(五) 時數：[2026-11-06]田凱文(3.00) |
| 第 10 週 | MDP - Infinite Horizon MDP 2026-11-13(五) 時數：[2026-11-13]田凱文(3.00) |
| 第 11 週 | RL - Introduction to model-free method 2026-11-20(五) 時數：[2026-11-20]田凱文(3.00) |
| 第 12 週 | RL - Monte Carlo Method 2026-11-27(五) 時數：[2026-11-27]田凱文(3.00) |
| 第 13 週 | RL - Temporal Difference (TD) Learning 2026-12-04(五) 時數：[2026-12-04]田凱文(3.00) |
| 第 14 週 | RL - TD Learning 2026-12-11(五) 時數：[2026-12-11]田凱文(3.00) |
| 第 15 週 | RL - Advanced Topics 2026-12-18(五) 時數：[2026-12-18]田凱文(3.00) |
| 第 16 週 | Final Project Presentation 2026-12-25(五) 時數：[2026-12-25]田凱文(3.00) |
• Ross, Sheldon M. (2014). Introduction to probability models. Academic press. • Warren B. Powell (2022). Reinforcement Learning and Stochastic Optimization: A unified framework for sequential decisions, John Wiley and Sons, Hoboken (free online) • Puterman, M. L. (2014). Markov decision processes: discrete stochastic dynamic programming. John Wiley & Sons. • Sutton R. & Barto A. (2020). Reinforcement Learning: An Introduction (2nd Edition). Cambridge: The MIT Press. (free online)
- 地點
- MB512
- 時間
- Mon. 12:00 – 14:00 (or by appointment)
- 聯絡方式
- kaiwen.tien@nycu.edu.tw