| Relevance of Course Objectives and Core Learning Outcomes(%) |
Teaching and Assessment Methods for Course Objectives |
| Course Objectives |
Competency Indicators |
Ratio(%) |
Teaching Methods |
Assessment Methods |
學生修課後能夠:
理解 AI 模型不可解釋性的本質
掌握模型層與系統層的 xAI 方法
評估解釋對使用者信任與決策的影響
結合 LLM 與 xAI 建構可解釋 AI 系統 |
| 1.Professional Knowledge and Practical Application |
| 2.Independent Analysis |
| 3.Innovative Research |
| 4.Leadership、communication and teamwork |
| 5.Social Responsibilities and Global Vision |
|
|
| topic Discussion / Production |
| Networking / Distance Education |
| Exercises |
| Discussion |
| Lecturing |
|
| Written Presentation |
| Attendance |
| Oral Presentation |
| Assignment |
| Internship |
| Other |
|
| Course Content and Homework/Schedule/Tests Schedule |
| Week |
Course Content |
| Week 1 |
為什麼 AI 需要解釋?
黑盒模型與可解釋性問題 |
| Week 2 |
傳統 xAI 方法(SHAP、LIME 概念)
LIME(局部線性模型)、SHAP(特徵貢獻度) |
| Week 3 |
深度模型的可解釋性
Grad-CAM(臉部情緒解釋) |
| Week 4 |
Integrated Gradients(文本解釋)
可解釋性評估指標 |
| Week 5 |
LLM
LLM 的可解釋性挑戰 |
| Week 6 |
LLM 作為解釋器(LLM-as-Explainer) |
| Week 7 |
Vibe Coding:xAI 工具實作 |
| Week 8 |
期中專題 |
| Week 9 |
資料偏誤分析&特徵偏誤(Feature Bias)
|
| Week 10 |
模型不確定性(Uncertainty)
可信度分析與模型錯誤來源 |
| Week 11 |
多模態 xAI
多模態衝突(face vs text)特徵分析 |
| Week 12 |
使用者導向的解釋設計
解釋品質、信任與決策 |
| Week 13 |
xAI 與 Human-in-the-loop
xAI 研究設計與評估方法 |
| Week 14 |
xAI 在產業決策的應用 |
| Week 15 |
xAI 研究設計與評估方法 |
| Week 16 |
期末專題 |
self-directed learning |
   03.Preparing presentations or reports related to industry and academia.
|
|
| Evaluation |
• 平時作業與實作:30%
• 期中專題提案:20%
• 期末專題成果:40%
• 課堂參與與討論:10%
|
| Textbook & other References |
|
| Teaching Aids & Teacher's Website |
| iLearning |
| Office Hours |
| 星期三下午四點到六點 |
| Sustainable Development Goals, SDGs(Link URL) |
| 04.Quality Education | include experience courses:N |
|