COMP3516 Data analytics for IoT / Spring 2025
Course Description
COMP3516 HKU Data analytics for IoT
Welcome to COMP3516@HKU!
This is the non-official site of HKU COMP3516 2025 Spring. This website is maintained at best effort. Enrolled HKU students should join Moodle page. In the event of any discrepancies between this page and Moodle, the Moodle page shall prevail.
This course introduces basic concepts, technologies, and applications of the Internet of Things (IoT), with a focus on data analytics. The course covers a range of enabling techniques in sensing, computing, analytics, learning for IoT and connects them to exciting applications in smart homes, healthcare, security, etc. The lectures cover the pipeline of data generation, data acquisition, data transportation, data analysis and learning, and data applications, with various topics from the fundamentals (e.g., signal processing, statistical analysis, machine learning) to real-world systems. Billions of things are connected today, and this course helps students to understand how IoT will evolve into AIoT (Artificial Intelligence of Things).
Important Note: According to the University’s guidelines, all lectures and tutorials will be delivered f2f in teaching venues. Lectures may be recorded, at best-effort, and video recordings will be uploaded to Moodle after the lectures.
| Lecture | Time/Day | Venues |
|---|---|---|
| Lecture-01 | 12:30 PM - 13:20 PM, Tue | CBA |
| Lecture-02 | 12:30 PM - 14:20 PM, Fri | CBA |
Grading
Note: The assignments are tentative and subject to changes. More details to be announced.
| Component | Weight | Description |
|---|---|---|
| Labs | 20 | 4 labs in total |
| Problem Sets | 10 | Two written problem sets |
| Project | 30 | Group project |
| Bonus | 1 | The Takeaway Talk (TTT)* |
| Final Exam | 40 |
*TTT: We intend to design and develop this course as a communication-intensive course. As part of it, we have designed a special component named “3-Minutes Takeaway Talk”. By the end of each lecture, I will reserve three minutes and invite one student to give a summary of the takeaway messages for the lecture, in the format of a Chalk Talk in front of the class. Every participant will be awarded a 1-point bonus. The participants will be selected (randomly, or at least as randomly as possible) by me at the end of the lecture. If you would like to participate, please sign up with your name first.
Besides TTT, as you can tell, the course also features intensive written communications (technical report, peer review, etc) and oral communications (presentation), plus wireless communications certainly:-P
Office Hour
| Instructor | Tutor | |
|---|---|---|
| Name | Dr. Chenshu WU | Mr. Sheng LYU |
| Office | Office: CB315B | HW101M |
| chenshu@cs.hku.hk | shenglyu@connect.hku.hk | |
| Office Hour | Tue 9:00 am-10:30 am Fri 9:00 am to 10:30 am By appointment: Appointment |
Thur 12:30 PM - 13:30 PM (f2f) Mon 9:30 AM - 10:30 AM (Online) By appointment: Appointment |
| Tel | 2241 5726 | 9476 5141 (Not preferred) |
Late Policy
Each assignment has a due date and a due time, which will be posted on the course web page. For assignments, we will use a “slip days” system to give you some flexibility with the assignment deadlines.
Each student starts the semester with five (5) slip days, which can be used to push back assignment deadlines. Slip days work as follows:
Pushing an assignment deadline back by one day (24 hours) costs one slip day. An assignment deadline can be pushed back at most three days. Fractional slip days are not allowed; e.g., it is not possible to use part of a slip day to push a deadline back by six hours. Slip days are not transferable from one student to another. By default, the slip days will be applied automatically* as needed according to the exact timestamp of submission. A FREE late window of 30 minutes is allowed to accommodate potential technical issues upon submission. In other words, submissions made within 30 minutes after the deadline need no slip days to cover them. All the thresholds will be observed strictly. For example, a submission that is late for 1801 seconds (i.e., 30 minutes plus 1 second) will use one slip day.
Assignments that are submitted late (with not enough slip days to cover them or more than 3 days) will still be accepted, but deducted to p = 0.99^t, where t is the uncovered late time rounded up in hours.
Let D be the deadline, and T your submission timestamp. The number of slip days that will be used for a submission, noted as SD, is calculated as follows:
if (T - D) <= 30 minutes, SD = 0;
else if (T - D) <= 24 hours, SD = 1;
else if (T - D) <= 48 hours, SD = 2;
else if (T - D) <= 72 hours, SD = 3;
else mark *= p.
These slip days are not for “procrastination.” They are designed to give you more breathing room in case some (usually unexpected and mostly unwanted) little things come up during the semester: short illnesses, injuries, visiting family or friends, a burst of work in other classes, etc. Use them wisely.
Review Window of Assignment Grading
Review of Assignment Grading (for any assignments) will ONLY be accepted within TWO weeks after the grading is released. Requests to review after two weeks will be ignored automatically without a reply.
