Unlock the power of data with mathematical precision.
Course Information
- Instructor: Shen-Ning Tung (tung@math.nthu.edu.tw)
- Lecture Time: Wednesdays, 3:30 PM – 6:00 PM
- Office Hours: Available for questions until 7:00 PM after each lecture (may leave early if no one is in the classroom).
Evaluation
- Weekly Problem Sets (20%)
- Midterm Report (30%)
- Final Project (50%)
Communication
- Primary Platform: All course communication and announcements will be made on the course Discord channel.
- Questions and Discussions: Please use public Discord posts for questions about course material. This fosters collaborative learning and allows instructors to address queries efficiently.
- Private Matters: For private discussions or questions, send a direct message to the instructors on Discord.
- Submissions: All notes and reports should be submitted via HackMD notes.
Course Project
The course project is an opportunity to delve deeper into the mathematical foundations of data science. You can work individually or in a group. Choose one of the following project types:
- Theoretical Deep Dive: Select a topic with a strong theoretical basis in data science and provide a comprehensive exploration, elucidating its key concepts, principles, and mathematical underpinnings.
- Algorithm in Action: Choose a data science algorithm, implement it, and apply it to a real-world dataset. Showcase its practical utility and interpret the results.
- Theory Meets Practice: Bridge the gap between theory and application. Introduce a topic or algorithm, explain its theoretical foundations, and then demonstrate its relevance and effectiveness by implementing it and analyzing its performance on real-world data.
Important: Please discuss your project topic with the instructor and finalize it by the end of September.
References
- Textbook: Foundations of Data Science by Avrim Blum, John Hopcroft, and Ravindran Kannan
- Probability and Statistics: All of Statistics by Larry Wasserman
- Linear Algebra: Linear Algebra and Learning from Data by Gilbert Strang
- Recommended: Mathematical Foundations of Data Sciences
Schedule
| Date | Lecture | Notes |
|---|---|---|
| 9/3 | Introduction to Data Science | Note |
| 9/10 | High-Dimensional Space | Note |
| 9/17 | Best-Fit Subspaces and SVD | Note |
| 9/24 | Random Walks and Markov Chains | Note |
| 10/2 | Typhoon Day Off | No class |
| 10/9 | Signal Processing | Note |
| 10/16 | Machine Learning | Note |
| 10/23 | Signature | Note |
| 10/30 | Algorithms for Massive Data | Note |
| 11/6 | Clustering | Note |
| 11/13 | Reinforcement Learning | Note |
| 11/20 | Network | Note |
| 11/27 | Optimization | Ref.1 · Ref.2 · Ref.3 |
| 12/4 | Operations Research | Note |
| 12/11 | Decentralized Finance | Ref.1 · Ref.2 |
Student Projects
- Apply Q-learning on McCall Search Model — Ya-Zhu Yang
- Reinforcement Learning for Snake — Felix Uhl
- Streamlining Time Series Analysis with sktime — Ren-Shu Yang
- Neural Network for Phase Transitions in the Ising Model — Hao-Yang Yen
- PySR: A Modern Symbolic Regression Method — YuanLong Chan
- Graph Partitioning and Community Detection — Jun-Zhi Wang