Link Search Menu Expand Document (external link)

DDA4210, Spring 2023

Course Materials

The course materials can be found in HERE

Online communication channels

  1. This website: a collection of useful resources, links, announcements, and course materials, etc.
  2. WeChat group: daily discussion and announcements, etc.
  3. Blackboard: The main channel for posting announcements, releasing and submitting assignments.

Logistics

Prerequisite: CSC4020/DDA3020/DDA2020/FTE4560

Session L02

Time: 01/04/2023-05/09/2023 Exam Week: (05/12-05/19)

Instructor: Tongxin Li litongxin@cuhk.edu.cn

Course Schedule: Mon & Wed 3:30PM - 4:50PM

Venue: Zhi Xin Bldg 109

Office Hours: Wed 5:00PM - 6:00PM

TA Office Hours: Tutorial schedule

Zoom Link:

Lectures (L02)

 https://caltech.zoom.us/j/87842855460?pwd=NkFUTlFZb1IxSUc4d2FISU0zNCtadz09 
 Meeting ID: 878 4285 5460  
 Code: 262095 

Tutorials

 https://cuhk-edu-cn.zoom.us/j/98464515700?pwd=UWZRZnUyZWxUaHh4RWtEbFFXYS9ydz09 
 Meeting ID: 984 6451 5700  
 Code: 123456 

Session L01

Instructor: Jicong Fan fanjicong@cuhk.edu.cn

Course Schedule: Mon & Wed 1:30PM - 2:50PM

Venue: Zhi Xin Bldg 109

Office Hours: Tuesday 4:00-5:30pm

TA Office Hours: Tutorial schedule

Zoom Link:

 Meeting ID: 214 182 2367  
 Code: N/A 

Course Information

Table of contents

  1. DDA4210, Spring 2023
    1. Course Materials
    2. Online communication channels
    3. Logistics
      1. Prerequisite: CSC4020/DDA3020/DDA2020/FTE4560
      2. Session L02
      3. Session L01
    4. Course Description
    5. Grading Scheme
    6. Policy for assignments, projects, and exams
      1. Assignments
      2. Projects
      3. Exams
      4. Academic Integrity
    7. Regrade requests
    8. Suggestions
    9. General Course Policies:
    10. Attendance requirement

Course Description

This course introduces advanced theory, algorithms, and applications of machine learning. Topics include: advanced ensemble learning methods; learning theory; advanced applications such as recommendation system; nonlinear dimension reduction, denoising, spectral clustering; generative models such as VAE and GAN; semi-supervised learning and GNN; causal inference, causal discovery, graphical and structural models, DAGs; attributes in trustworthy machine learning, interpretability and explainability, biases, fairness and safety, fair causal reasoning, adversarial machine learning, NN robustness and applications. This course requires students to have intermediate mathematical and programming backgrounds in machine learning and deep learning.

Grading Scheme

  • Assignments and CTE: 30%
  • Projects: 40%
  • Final exam: 30%

(There maybe some in-class quizzes, contributing a random portion from 0% to 5% to the final score.)

Policy for assignments, projects, and exams

Assignments

  • There will be three assignments in total. Problem sets will be assigned approximately tri-weekly. Your solutions should be submitted as a single PDF file to Blackboard by 10pm Beijing Time (or a particularly specified time) on the due date, or earlier. More details and instructions about the homework submission will be provided in the assignments.
  • Late homework policy: Late submissions will not be considered.
  • Collaboration is not allowed for all problems on your homework sheet. The Honor Code is taken very seriously in this course and we have no tolerance for behavior that falls outside our boundaries for acceptable conduct. Please do your part in maintaining a community where academic work is done with a high standard of integrity.

  • Assignments and due dates will be announced on the course website. There will be reminders of the coming deadlines at the beginning of each week in our WeChat Group and/or via email.

Projects

  • There will be two projects. The first one is a mini-competition and the second one is a group project, which will be evaluated based on the final report and code (25%) and in-class group presentation (75%=15%peer+25%TA+35%instructor).
  • The first project and second project will contribute to 15% and 25% of the total score respectively.
  • Students can form a group up to 4 people for the course project. The grading scheme will be the same regardless of the size of the group, as long as it is less than or equal to 4.
  • Students are allowed to form groups across different sessions (L01 and L02).

Exams

  • Exams will be in-person or online, depending on the situations. The specific time and form will be announced later.

  • Absences: Make-up exams will only be allowed under extraordinary and unavoidable circumstances. Appropriate documentation verifying the absence may be required; in cases of illness, you will be asked to have official documentation from a physician-reviewed and verified by the Dean of the school. Make-up exam arrangements will be made based on a case-by-case basis.

Academic Integrity

  • Plagiarism will be dealt with severity. See “Academic Integrity” below for more details.

  • Grading clarifications (in assignments as well as exams) should be resolved within a week from the date when the graded submission is returned. No clarification applications will be considered after a week. Bring your clarification requirements to the TA’s office hour.

Regrade requests

Credit for work will be recorded only as reported by the TA in Blackboard. It is your responsibility to make sure that your work has been properly recorded in Blackboard.

If you need to request a regrade for an assignment, please adhere to the following policy:

  • If there was a clerical or math mistake in calculating or recording your grade, please go to office hours of the TA who graded it so that it can be fixed.
  • If you lost points for something you did correctly, e.g. the grader said “-2 points for not doing X” and you actually did do that (or something similar), please speak with the TA who graded it during their office hours. It is preferable to discuss these things in-person than over email. However, if you can’t attend the grader’s office hours, then email them to try to arrange another time to meet.
  • Likewise, if you lost points for something and do not understand the grader’s comments, please speak with the TA who graded it during their office hours if possible.
  • For any other issues, please contact the instructor. This may be things like “I didn’t realize we had to do X,” “I misunderstood this part of the assignment,” etc. This isn’t to say that you’ll necessarily get points back for your misunderstanding, but issues such as these should be discussed with the instructor.

Regrade requests must be made within one week of the score being posted in blackboard. Only regrades related to administrative mistakes (e.g., miscalculating the score or entering it incorrectly) made after the one-week period are likely to be considered.

Suggestions

  • Frequently check WeChat group and Blackboard for announcements!

  • Email TAs for homework-related questions (can either cc or not cc the instructor).

  • Email the instructor for other questions.

  • Check Blackboard and this website; e.g., once every 2-3 days.

  • Feel free to ask questions in the WeChat group. No question is stupid!

  • You can email the instructor for most questions regarding course contents or logistics, but the instructor may or may not be able to respond in time (though the instructor often replies within 24 hours). It is recommended to ask questions in the Wechat group, or email the TAs (you may consider cc’ing the instructor). If it is urgent or you really want to hear from the instructor, you can add “[DDA4210 Urgent]” or “[DDA4210 Need Response]” to the title, so that the instructor will pay attention to the email. · You are encouraged to discuss lecture materials and practice problems with each other; but not the homework problems.

General Course Policies:

Academic Integrity: Academic dishonesty may result in a failing grade (i.e. F). Every student is expected to review and abide by the Academic Integrity Policy of CUHK-Shenzhen. Ignorance will not be allowed as an excuse for any academic dishonesty. Do not hesitate to ask me if you are ever in doubt about what constitutes plagiarism, cheating, or any other breach of academic integrity.

Attendance requirement

You are expected to attend the course in-person or watch all course videos on time unless you are unable to come to the campus (don’t wait until the last minute of the assignment deadline). We use quizzes help you keep the pace. Important course announcements will be made in Blackboard and this website or via email; you are responsible for being aware of these announcements.