Welcome to Graph Filters for Processing and Learning Data over Graphs (2021)


Description: Data generated by networks are almost everywhere such as data generated by social, biological, or sensor networks, to name just a few. Network data are particular because they are complex and irregular, therefore, do not lend themselves to standard processing tools. Recent years have seen a surge of interest to extend signal processing and machine learning concepts to network data. In this course, we will cover the fundamental concepts to process and learn from data over networks. In particular, we will discuss concepts from graph signal processing, graph neural networks, and detail their role for a series of network data problems such authorship attribution, recommender systems, and multivariate time-series forecasting.

Specific topics include: the graph Fourier transform, regularisation over graphs, graph filters. models for time-varying processes over graphs, graph neural networks, edge varying graph neural networks (EdgeNets), graph-adaptive activation functions, graph-time convolutional networks, and graph dropout.

After this course, you will have knowledge on:

1. The underlying mathematical principles about processing and learning from data over networks.
2. Graph neural networks and different architectures.

3. Applications of network data processing tasks.

The course is organised in the form of lectures with working group exercises and interactive discussions. Students attending this course need to have basic knowledge of linear algebra, machine learning, and deep learning. Evaluation methods include a exam of 30 minutes.



Organizer: Associate Professor Chenjuan Guo - cguo@cs.aau.dk

Lecturers: Assistant Professor Elvin Isufi - E.Isufi-1@tudelft.nl

ECTS: 2.0

Time: 31 May - 01 June 2021

Place: ONLINE

Zip code: 
9220

City: Aalborg

Number of seats: 60

Deadline: 10 May 2021


Important information concerning PhD courses: We have over some time experienced problems with no-show for both project and general courses. It has now reached a point where we are forced to take action. Therefore, the Doctoral School has decided to introduce a no-show fee of DKK 3.000 for each course where the student does not show up. Cancellations are accepted no later than 2 weeks before start of the course. Registered illness is of course an acceptable reason for not showing up on those days. Furthermore, all courses open for registration approximately four months before start. This can hopefully also provide new students a chance to register for courses during the year. We look forward to your registrations.