COMP9312 Notices

Good Luck

Posted by Dong Wen Thursday 01 September 2022, 06:27:03 PM.

Dear students,

I think you have already received your final marks today (excepted some supp exam students). Hope you are happy with that. This should be the last email from COMP9312 22T2. I think you must have learned something and known some interesting people more or less in this short 10-week term. I am also pleased and enjoy my teaching experience in 9312_22T2. Wish you all the best of luck.

Regards,
Dong

Exam end soon

Posted by Dong Wen Thursday 25 August 2022, 04:48:14 PM.

The exam will end in around 10 mins. Please make sure you have submitted your answer to Moodle before 5pm.

Exam paper

Posted by Dong Wen Thursday 25 August 2022, 12:49:18 PM.

The entry link of exam paper has been released in your Moodle final exam session. Please do not share the entry link with others. Let us know if you have any questions during the exam.

Final Exam

Posted by Dong Wen Wednesday 24 August 2022, 12:48:10 PM.

Dear students,

Our final exam will be in 13:00--17:00 Thursday 25th Aug . We will release the entry link of the exam paper (via notice email and webcms) at around 12:50 so you will have some time to read general instructions and review the exam paper. If you have any questions or urgent issues during the exam, just send us an email, and we will response ASAP. All exam answers are required to written into a single pdf file . At least, you will need a mail reader to receive any updates, a browser to view the exam paper, and a rich text editor that can export a pdf file.

Our final response rate for MyExperience survey is over 50%. As we mentioned earlier, we will apply 1 mark bonus for every student. Good luck~

Ass2 Mark Release

Posted by Dong Wen Thursday 18 August 2022, 02:30:13 PM.

Dear students,

Good afternoon. Assignment 2 marks have been released on Moodle. Let us know if you have any questions.

Dong

Project Marks

Posted by Dong Wen Thursday 11 August 2022, 07:56:53 PM.

Dear students,

Project marks have been released on Moodle. Contact us for any marking questions.

Regards,

Dong

Sample questions and Survey

Posted by Dong Wen Sunday 07 August 2022, 07:23:45 PM.

Dear students,

We have provided several sample questions for the final exam. You can find it via the following link. Let us know if you have any question.

https://cgi.cse.unsw.edu.au/~cs9312/22T2/sample_exam/

The response rate of MyExperience survey is still low now (29%), and the school is unhappy with that. Please do the course evaluation to make them happier. For a sweetner, we will give everyone a 1 mark bonus on the Exam if the response rate hits 50% before 11th Aug.

MyExperience survey and final exam

Posted by Dong Wen Tuesday 02 August 2022, 01:24:47 PM.

Dear students,

Welcome the last lecture day of 9312. We will review the course and talk about the final exam in today's lecture. Feel free to ask questions. Below are some key dates and tips.

Final Exam. You should have already known that the final exam will be on 1--5pm 25th Aug (I know it is very late... but it is arranged by the school by considering exam collisions of each student). Several sample questions will be provided later this week.

Exam consultation. Directly send us an email, and we will arrange a private meeting (online or F2F) for each student.

Survey. MyExperience survey has opened. You can find the entry on Moodle or directly use this link . The response rate for MyExperience has been pretty low so far (only 8%). MyExp team is crying about this (sob, sob~). Please fill the survey at your convenience. The deadline is 11th Aug.

Regards,

Dong

Assignment 2 Released

Posted by Michael Ruisi Yu Tuesday 26 July 2022, 06:31:19 PM.

Dear students,

Note that Assignment 2 has been released: https://cgi.cse.unsw.edu.au/~cs9312/22T2/assignment/ass2/

Note: You should be able to complete the majority of the assignment 2, except for one question on color-refinement (WL Kernel) .
The related slides for that question are now available on webcms3 (see slides on Topic 9). I will cover that on Tuesday Wk10, after the guest lecture on Monday.

The deadline is at 9pm Friday Week 10

On next Monday, we are lucky to invite Dr. HanCheng Wang who specializes on machine learning on graphs to talk about advanced topics relating to GNNs, we will release details about his talk as soon as it is finalized.

Best,
Michael

Project Extension + Update to Project Q2

Posted by Michael Ruisi Yu Tuesday 19 July 2022, 12:37:26 PM.

Good Afternoon Students,

Last night, we have found Q2 to contain some minor irregularities in our formulation of the GCN layer definition through the Graph Laplacian.
To clear up remaining confusion, we have changed the definition of GCN layer. It now matches exactly the definition that proposed in the paper by Thomas N. Kipf and Max Welling https://openreview.net/pdf?id=SJU4ayYgl . (See Equation 2 )

Fortunately If you've already done the project, this change should not introduce any additional work other than a change in the layer definition. Nevertheless, since we've changed the definition, we will also provide everyone with more time for you to account for this adjustment.

Since we don't want to set a deadline for the weekend, we will extend the deadline to Monday evening next week ( 9pm of July 25th ). In total, you now have 3 more days to make the change and further explore/refine your work. We hope this will minimize any inconveniences with completing the project on time.

Best,
Michael

A change in the week7 tutorial material

Posted by Michael Ruisi Yu Thursday 14 July 2022, 12:05:55 AM.

Evening students,

Since Jupyter Notebook now uses a newer version of PyTorch, a minor edit has been made to the week 7 tutorial material to adjust to this recent change.

Specifically, we updated code cell [16] relating to installing the relevant binaries.
The command now specifies 1.12.0+cu113 to match the specific version of PyTorch supported by Jupyter, where it was previously 1.11.0+cu113.

This fix should resolve any troubles you might have experienced today on Jupyter Notebook.

Best,
Michael

Week 7

Posted by Dong Wen Monday 11 July 2022, 01:21:10 PM.

Hope you had a great weekend. Starting from this week (Week 7), Dr. Michael Yu will introduce tropics about learning-based techniques in graph processing. The slides for today's lecture have been uploaded to Webcms.

https://webcms3.cse.unsw.edu.au/COMP9312/22T2/resources/74998

Ass 1 Results

Posted by Dong Wen Thursday 07 July 2022, 09:44:57 PM.

Hi all,

Marks for Assignment 1 have been released on Moodle . Some feedback comments and the marking tutor's email are included. If there is any marking question or mistake, send an email to the corresponding marking tutor and cc me (dong.wen@unsw.edu.au). We will double check and fix it. Thanks.

Regards,

Dong

Project Q1 Update

Posted by Dong Wen Thursday 30 June 2022, 04:12:36 PM.

Hi,

Some students reported that the current test dataset (citation.graph) in Q1 is relatively large, which is not convenient to test/debug algorithms. Therefore, we have replaced citation.graph by three small datasets ( cora , map_BJ_part , and map_NY_part ). Details can be found in Section 3.1 of Q1.ipynb. Update records can be found on the update page .

Tips: you can just copy the code and download all datasets to your familiar local environment/editors, which may be more convenient to develop and debug.

There are no lectures and tutorials next week (week 6). Enjoy your holiday but feel free to let me know if you have any question.

Regards,

Dong

COMP9312 Project

Posted by Dong Wen Monday 27 June 2022, 03:43:56 PM.

We have released the project on Webcms. The deadline is 9pm Friday 22 July . Let us know if you have any question.

https://webcms3.cse.unsw.edu.au/COMP9312/22T2/resources/75063

Census Date

Posted by Dong Wen Friday 24 June 2022, 11:58:40 AM.

This is a gentle reminder for census this weekend. If you remain enrolled in courses after 11:59 on Sunday 26 June, you'll be charged a fee for each enrolled course.

Assignment 1 Due and Guest Lecture

Posted by Dong Wen Wednesday 22 June 2022, 12:08:18 PM.

Ass1. This is a gentle reminder that the assignment 1 is due by 9pm Friday (24th Jun).

Guest Lecture. On Monday next week, we will invite a research fellow Dr. Li from UTS to give a guest lecture. The title of his talk will be "Measuring Graphs at Billion-Scale". His research covers a set of state-or-the-art solution for several fundamental problems like short path and reachability, which are closely related to the course. You are also welcome to ask questions after his talk~

Code for week 2 & No lecture on Monday Week 3

Posted by Dong Wen Tuesday 07 June 2022, 06:44:30 PM.

Python code for the online lectures of week 2 has been uploaded to webcms. I have added the implementation for the disjoint set data structure.

https://webcms3.cse.unsw.edu.au/COMP9312/22T2/resources/74998

In our tutorials, you may face up the same question but with the different input formats, different function interfaces, which are common in real cases. You can easily handle any scenario (even any programming language in practice) if you fully understand all these fundamental algorithms.

The first part of the course is somehow an algorithm course, but in addition to the theoretical stuffs, we aim to help you implement each algorithm in a correct way and make your code extremely fast.

Enjoy~

Assignment 1

Posted by Dong Wen Monday 06 June 2022, 01:26:44 PM.

Welcome to Week 2. We have released the Assignment 1 on webcms.

https://webcms3.cse.unsw.edu.au/COMP9312/22T2/resources/75063

The deadline is at 9pm Friday Week 4 . Let us know if there is any typo/error on the Ass1 paper.

Lecture Session Entry

Posted by Dong Wen Monday 30 May 2022, 01:26:51 PM.

Lecture slides and the entry link for week 1 Monday session have been uploaded to webcms. We will open the session at around 3:50PM.

https://webcms3.cse.unsw.edu.au/COMP9312/22T2/resources/74998

COMP9312 Welcome

Posted by Dong Wen Wednesday 25 May 2022, 01:36:53 PM.

Hi, all, and welcome to COMP9312 Data Analytics for Graphs!

We aim to introduce some basic algorithms for big graph processing given the prevalence of graph data in many areas. The course will also summarize some recent research studies which may never be discovered in any textbook.

As you can see, we still run a significant majority of course works online, including lectures, and all assessments. One tutorial session is offline and provides you a chance of F2F communication with your tutors. We manage the course with WebCMS3 ( https://webcms3.cse.unsw.edu.au/COMP9312/22T2/ ) and Moodle . All announcements, lecture notes, tutorial notes, assignment spec, etc. are going to be posted on WebCMS3. Moodle is used to provide access to online live lectures and lecture recordings.

We will deliver lectures live online at the scheduled lecture time 16:00 – 18:00 Monday and 16:00 – 18:00 Tuesday. We understand that some students are overseas with different time zones and hope this will not cause too much inconvenience to you. Lectures will be delivered and also recorded in Blackboard Collaborate in Moodle. There are two ways to access the lecture:

1) You can access it directly from Moodle ( https://moodle.telt.unsw.edu.au ), find our course, click "lectures and recordings", click Menu (in up left corner), click "Sessions" to join a live lecture or "Recordings" to access the recorded lecture.

2) We will provide a copy of the link to access the lecture (live and recording) in the WebCMS Lectures Tab.

Tutorials (online+offline) will start from Week 2. Consultations will follow Tuesday lecture sessions (18:00 - 19:00). You are also welcome to send us an email for private help session.

Before week 1, you can have a look at the course outline on WebCMS3. COMP9312 is a new course (2nd term). We are still continuously improving the course materials and optimizing the syllabus. Your feedbacks are important for us to improve the course. Feel free to send me an email and share your suggestions for the course.

Look forward to meeting you soon~

Dong Wen

Lecturer-in-Charge