Seminar Geometry and Topology in Deep Learning

With machine learning making increasing use of foundational principles from geometry and topology, this seminar provides a glimpse into the state of the art. We will discuss architectures and use cases of models leveraging both geometry and topology to address complex tasks, including graph classification, graph regression and general shape analysis.

Details

Code	63612
Type	Seminar
ECTS	5
Site	Fribourg
Track(s)	T6 – Data Science
Semester	A2024

Teaching

Learning Outcomes	The primary goal of this seminar is for students to learn how to approach cutting-edge literature in machine learning research. Students will learn how to read and analyse papers, followed by a brief report on a specific body of work. Moreover, students will be given the opportunity to (re)implement methods described in the paper, using a programming language of their choice. Successful participation in this seminar will enable students to undertake small guided research on their own.
Lecturer(s)	Bastian Grossenbacher-Rieck
Language	english
Course Page	The course page in ILIAS can be found at https://ilias.unibe.ch/goto_ilias3_unibe_crs_3102208.html.

Schedules and Rooms

Period	On Appointment
Location	UniFR, PER21

Evaluation

Evaluation type

continuous evaluation

Additional information

Comment

First Lecture The first lecture will be announced later.