人工智慧導論
Introduction to Artificial Intelligence
| 節 | 週三 |
|---|---|
5 13:20–14:10 | 人工智慧導論 A202（光復） 2 節連堂 |
6 14:20–15:10 |
* 根據陽明交大上課時間表所列
Rapid advances in artificial intelligence (AI) in recent are expected to transform how people work, communicate, and live fundamentally. To enable the general public to understand the true nature of AI for the future, this course begins with the historical development of artificial intelligence and provides an accessible, in-depth introduction to the principles, concepts, and application domains of current AI technologies. Through intuitive explanations and visualized learning tools, students will gain a concrete understanding of how modern machine/deep learning techniques operate. Besides, in the final part of the course, students will be guided to use cutting-edge Harness / Agentic Engineering techniques to turn their own ideas into realuty with the help of AI tools. Through this hands-on process, students will develop a thorough understanding of both the power and the limitations of contemporary artificial intelligence.
Students need to have basic computer operation skills and are encouraged to bring laptops, tablets, or other devices capable of opening web pages to class. The course relates to foundational concepts in calculus, linear algebra, and probability and statistics, but prior knowledge in these areas is not required, as they will be taught during the course.
無備註
Explain various AI technologies in a clear and accessible way, accompanied by Excel, visualizations, and online tools to help students experience the principles covered in class. Finally, teach how to effectively communicate with current AI tools to produce deliverables, and shape the communication process into a skill to facilitate Agent execution. Through brainstorming, have students think about problems in daily life that need to be solved along with possible solutions, and ultimately bring these to fruition using the methods taught in the course.
隨堂測驗20% A brief quiz will be conducted at the end of each class session, with approximately 10–12 quizzes in total. 作業20% Hands-on activities using designated tools to explore the principles of machine learning, deep learning, and reinforcement learning, with approximately 3–4 activities in total. 報告15% Conducted in groups: students will attempt to induce hallucinations in an LLM, explain the causes of these hallucinations, and use prompting strategies to correct them. Each group will submit a report documenting the process. 期末考15% A written exam covering the course content. 討論5% Conducted in groups: students will use brainstorming techniques to discuss the requirements of their final project and explore possible digital solutions. 展演25% Conducted in groups: students will collaborate with an LLM to implement their final project ideas, present the process and methods used, and demonstrate their outcomes.
History of Artificial Intelligence
1. The Rise and Fall of AI 2. BDI Agent & Ontology 3. Genetic Algorithm 4. Expert System
- 講授：
- 1
- 示範：
- 0.5
- 實作：
- 0.5
The Foundation of Machine Learning
1. Kolb's Experiential Learning Cycle 2. The Principle of Machine Learning
- 講授：
- 2
Classic Machine Learning Methodology
1. Decision Tree 2. Support Vector Machine 3. Ensemble Learning
- 講授：
- 2
- 示範：
- 0.5
- 實作：
- 0.5
The Principle of Deep Learning
1. Representation Learning 2. Multi-Layer Perceptron 3. Gradient Descent and Back Propagation
- 講授：
- 2
- 示範：
- 0.5
- 實作：
- 0.5
Various Techniques of Deep Learning
1. Convolutional Neural Network and Computer Vision 2. Recurrent Neural Network 3. Attention 4. Autoencoder 5. Transformer 6. Generative Adversarial Networks (GAN) 7. Diffusion Model
- 講授：
- 6
- 示範：
- 2
- 實作：
- 2
Reinforcement Learning
1. Markov Process 2. Q-Learning 3. Deep Reinforcement Learning
- 講授：
- 2
- 示範：
- 0.5
- 實作：
- 1
Large Language Model & Harness/Agentic Engineering
1. The principle of LLM 2. The power and limitation of LLM 3. Harnessing Your LLM 4. Building Your Own Agent
- 講授：
- 2
- 示範：
- 1
- 實作：
- 3
History of Artificial Intelligence
1. The Rise and Fall of AI 2. BDI Agent & Ontology 3. Genetic Algorithm 4. Expert System
- 講授：
- 1
- 示範：
- 0.5
- 實作：
- 0.5
| 週次 | 主題 |
|---|---|
| 第 1 週 | 單元主題： The History of Artificial Intelligence (1) The Rise and Fall of AI Introduction to BDI Agent & Ontology, Genetic Algorithm, and Expert System 2026-09-09(三) |
| 第 2 週 | 單元主題: The Foundation of Machine Learning (1) Kolb's Experiential Learning Cycle (2) The Principle of Machine Learning (3) Clustering Methods (Unsupervised Learning) Hands-on Practices in Hierarchical Clustering 2026-09-16(三) |
| 第 3 週 | 單元主題: Classic Machine Learning Methodology (1) Introduction to Supervised Learning (2) Decision Tree (3) Support Vector Machine (4) Ensemble Learning Hands-on Practices in Decision Tree 2026-09-23(三) |
| 第 4 週 | 單元主題: The Principle of Deep Learning (1) Representation Learning Multi-Layer Perceptron 2026-09-30(三) |
| 第 5 週 | 單元主題: Gradient Descent and Back Propagation Hands-on Practices in Gradient Descent and Back Propagation 2026-10-07(三) |
| 第 6 週 | 單元主題: Convolutional Neural Network and Computer Vision (1) How Machine Sees (2) What is Convolution (3) How CNN Works Gradient Descent and Back Propagation for CNN 2026-10-14(三) |
| 第 7 週 | 單元主題: Recurrent Neural Network and Attention (1) Why MLP doesn't Fit Sequence Data (2) The Principle of RNN (3) What is Attention Hands-on Practices in Attention 2026-10-21(三) |
| 第 8 週 | 單元主題: Attention is All you Need Attention and Transformer 2026-10-28(三) |
| 第 9 週 | 單元主題: Self-Supervised Learning (1) Autoencoder (2) Generative Adversarial Networks (GAN) (3) Diffusion Model 2026-11-04(三) |
| 第 10 週 | 單元主題: Reinforcement Learning (1) Markov Process (2) Value Function (3) Q-Learning Hands-on Practices in Value Function-based RL 2026-11-11(三) |
| 第 11 週 | 單元主題: Deep Reinforcement Learning (1) Deep Q-Learning (2) Monte-Carlo and Temporal Differences (3) Actor-Critic methods (4) Policy Gradient Reward Shaping and Imitation Learning 2026-11-18(三) |
| 第 12 週 | 單元主題: Large Language Model (1) What is LLM (2) How to use LLM properly (3) Why there is Hallucination Report writing: Triggering hallucinations in LLM and the solution of it. 2026-11-25(三) |
| 第 13 週 | 單元主題 Working with LLM (1) Harnessing an LLM (2) Making Skills (3) CLI and MCP for Agents Demonstration of Making a Web-based Application using LLM 2026-12-02(三) |
| 第 14 週 | 單元主題: Brain storming for Requirements if the Term Project 2026-12-09(三) |
| 第 15 週 | 單元主題: Final Exam 2026-12-16(三) |
| 第 16 週 | 單元主題: Demonstration of the Term Project 2026-12-23(三) |
教科書列表：Using instructor-developed course materials. 自製講義或課本之參考書目： https://www.youtube.com/@HungyiLeeNTU Professor Hung-yi Lee’s online video courses on various topics in artificial intelligence, National Taiwan University. https://ml-visualized.com/index.html Visualization of Machine/Deep Learning Mechanism
- 地點
- TBC
- 時間
- Tuesday 1pm - 3pm Wednesday 10am-12pm
- 聯絡方式
- TBC