Graph Machine Learning Foundations and Applications (AI60007) / Autumn,2024

News

Course webpage is up and running.
Moodle accounts have been created for everyone. You may visit Moodle.
Assignment-1 has been announced, please check your moodle accounts.
Groups have been formed and pdf is available at Assignment section.
Assignment-2 has been announced, please check your moodle accounts.
Assignment-1 marks have been released, please check your moodle accounts.
Three new tutorials have been uploaded, please check Hands-on section.
Assignment-2 marks have been released, please check your moodle accounts.
Assignment-3 has been announced, please check your moodle accounts.
Assignment-4 has been announced, please check your moodle accounts.

Course Description

Graphs have widely been used in representing problem scenarios in different domains. Subsequently, several computational models have been developed to process graph representations. Machine learning, like in many other domains, have been used to model different tasks on graphs including graph clustering, node classification, link prediction etc. Deep learning models though are proven to be very effective in domains like computer vision, text processing etc., cannot be directly applicable in irregular data representation like graphs. Subsequently, a new paradigm of machine learning (graph machine learning) has emerged. Right from its emergence, it has been widely adopted in diverse domains including social networks, protein-protein interaction networks, drug disease interactions etc. Along with its effectiveness on graph data, graph machine learning has also improved the state-of-art in the domains (e.g., computer vision, text processing) where traditional deep learning architectures have ruled. Considering the rapid adoption of graph machine learning and its effectiveness, this course is proposed to develop the theoretical foundation of graph machine learning models and their use in real world applications in diverse domains.

Course Logistics

Lecture Hours:
- Wednesday [11:00 AM to 11:55 AM]
- Thursday [12:00 PM to 12:55 PM]
- Friday [08:00 AM to 08:55 AM]
Classroom: Nalanda Classroom Complex - NR-213
L-T-P & Credits: 3-0-0 & 3 Credits

Grading Policy

Semester Examinations:
- Mid-Semester Examination: 25%
- End-Semester Examination: 35%
Assignment: 35%
Attendance: 5%

Assignment Plagiarism Policy

Code plagiarism will be checked on the designated blocks which the students have to code.
No penalty till 20% (Baseline).
Each 10% increase over baseline will cost 1 mark, Beyond 70% absolute -2 (negative).
Two groups having similar code will be penalized with the same amount. No arbitration will be done.
Late submission policy: One week window with a penalty of 3 marks.

Advisory: Using GenAI tools for solving assignment is discouraged as this will increase the probability of plagiarism with others who may choose the same path. [One can always consult TA's on their given emails for any further clarification (or) any help]

Assignment [Submission Portal and Groups]

You can submit the assignments by clicking on this submission portal link.

Moodle Course Groups

Assignment 1 due date: [23-08-2024]

Assignment 2 due date: [13-09-2024]

Assignment 3 due date: [17-10-2024]

Assignment 4 due date: [31-10-2024]

Prerequisites

Deep Learning: Foundations and Applications (AI61002) or Deep Learning (CS60010)

Honor Code

Academic integrity is very important for us. You are required to follow the honor code to maintain academic integrity.

Your solutions against assignments, tests must be entirely your own (Exception: You may collaborate if instructed by the faculty).
You may not share your solutions for the scheduled assignments and tests with your peers unless instructed by the faculty.

Previous Offerings

Autumn-22

Instructor

Plaban Kumar Bhowmick

Teaching Assistants

Animesh

ainimesh1@gmail.com

Trishita Mukherjee

trishitamukherjee77@gmail.com

Kishalay Das

kishalay.msit@gmail.com