[AI-announce] Courses of the research group "Logic and Formal Verification" in winter term 23/24

[Thomas Zeume]

Dear students,

Soon the winter term 2023/2024 will start and I would like to draw your 
attention to the courses offered by our research group.

## Bachelor lecture "Logik in der Informatik".

Lecture: Mondays 10:00 - 12:00, room HGB 10 (first lecture: 09.10.2023)
https://moodle.ruhr-uni-bochum.de/enrol/index.php?id=53920


## Master lecture "Computational complexity theory"
Lecture: Tuesdays and Thursdays, 10:00 - 12:00, Room MC 1.54 (first 
lecture: 10.10.2023)
https://moodle.ruhr-uni-bochum.de/enrol/index.php?id=53919


## Bachelor seminar "Perlen der theoretischen Informatik"
Registration has already closed.


## Master seminar "Learning and Formal Methods"
This seminar is at the interface of AI, formal methods and theoretical 
computer science.  It is a joint seminar with Prof. Dr. Daniel Neider 
(Verification and Formal Guarantees of Machine Learning) and Prof. Dr. 
Jean Jung (Knowledge-Based Systems) at TU Dortmund. See below for more 
information on content.

Dates:
- Introduction session: 11.10.2023; 10 am
- Talks: en bloc in the week 5.2.2024 - 9.2.2024

Registration: via email to me (thomas.zeume at rub.de)



I wish you a good start to the winter term!

Best wishes,
Thomas Zeume


Seminar "Learning and Formal Methods"

Abstract: In this seminar we will explore recent advances in  how (1) 
formal descriptions and models can be learned from data; and how (2) 
methods from theoretical computer science can be used to understand the 
power of models used in machine learning, such as (graph) neural networks.

Learning formal descriptions and models from data has many applications 
in areas such as formal verification, databases, artificial 
intelligence, and knowledge representation. A typical challenge in these 
areas is to design algorithms that learn logical formulas, queries, or 
specifications from user-provided  data.

On the other hand, theoretical computer science provides the means to 
understand the power of typical models used in machine learning. For 
example, the power of graph neural networks has recently been related to 
the computational power of classes of circuits and to the expressive 
power of logics.

In the seminar, we will survey the recent literature on these aspects. 
Most topics will be of rather theoretical nature (in the sense of 
theoretical computer science), but there will also be some more 
practical topics.

The seminar is aimed at Master students of Computer Science, ITS, and 
Mathematics.


-- 
Prof. Dr. Thomas Zeume
Logic and formal verification

Faculty of Computer Science
Ruhr University Bochum
Universitätsstraße 150
44801 Bochum

Phone: +49 (0)234 32-19609
Mail: thomas.zeume at rub.de