Jump to contentJump to search

Jannik Dunkelau

Contact Information

Heinrich Heine University Düsseldorf
Institut für Informatik
Universitätsstraße 1
D-40225 Düsseldorf
Room: 25.12.02.52

Phone: +49 211 81-12635
Email:

Research Interests

  • Integration of formal methods with machine learning and artificial intelligence
    • Machine learning for improvement of formal methods
    • Verification of artificial neural networks and machine learning systems
  • Symbolic and explainable AI
  • Fairness-aware Machine Learning

Past Courses

  • Overview of Artificial Intelligence (Seminar, Summer Term 2020)
  • Overview of Artificial Intelligence (Seminar, Summer Term 2019)
  • Safety-Critical Systems (Tutorial, Winter Term 2018)
  • Programming Project I (Tutorial, Summer Term 2018)

Concluded Theses

I supervised the following theses.

Name

Title

Type

Dominik Brandt Implementing General Game Playing in Prolog: Applying MCTS to Three Examples Bachelor Thesis
Tim Richter Autoencoding ProB Constraints for Classification Tasks Bachelor Thesis
Leo Baldus Backend Selection for ProB Based on Ranked Runtime Predictions Bachelor Thesis
Benedikt Jung Vergleich fairer Machine-Learning-Algorithmen Bachelor Thesis
Aycan Aytan The Preference Beyond: Applying Deep Forests to Predict Individual Reward Preferences From Reward-Locked ERPs Bachelor Thesis
Jessica Petrasch The Decision Does Not Fall Far from the Tree: Automatic Configuration of Predicate Solving Master Thesis

Open topics for bachelor and master theses can be found here (German).


Publications

2020

2019

2018

  • Machine Learning and AI Techniques for Automated Tool Selection for Formal Methods.
    In Proceedings of the PhD Symposium at iFM’18 on Formal Methods:Algorithms, Tools and Applications, Research Report, 483, University of Oslo, 2018.
Responsible for the content: