Research Focus

My current research focuses on the development of theoretically sound algorithms for sequential decision tasks with weakly supervised feedback, e.g., preference information or censored observations, with automated algorithm configuration being the main application area. My recent research ventures also include the study of fundamental properties for quantifying uncertainty as well as investigating research questions in cooperative game theory.

Research areas

Bandit Algorithms, Uncertainty Quantification, Preference-based Learning, and Explainable AI

Professional Activities

Editorial Board Member

  • Machine Learning Journal

Program Committees of Conferences

  • European Conference on Machine Learning/Principles and Practice of Knowledge Discovery in Databases (ECML PKDD)
  • Discovery Science (DS)
  • European Conference on Artificial Intelligence (ECAI)

Invited Reviewer for Journals

  • Metrika
  • Machine Learning Journal
  • Data Mining and Knowledge Discovery
  • Mathematics of Operations Research

Invited Reviewer for Conference

  • International Conference on Machine Learning (ICML)
  • International Joint Conferences on Artificial Intelligence (IJCAI)
  • Conference on Neural Information Processing Systems (NeurIPS)
  • International Conference on Artificial Intelligence and Statistics (AISTATS)
  • Conference on Uncertainty in Artificial Intelligence (UAI)

Conference Talks and Posters

Talks

  • On the fundamental flaw of quantifying epistemic uncertainty through loss minimisation, Workshop on Uncertainty in Machine Learning (WUML), Gent, Belgium, 2023
  • Pitfalls of epistemic uncertainty quantification through loss minimisation, Advances in Neural Information Processing Systems (NeurIPS), hybrid conference, New Orleans, USA, 2022
  • Stochastic contextual dueling bandits under linear stochastic transitivity models, International Conference on Machine Learning (ICML), hybrid conference, Baltimore, USA, 2022
  • On the difficulty of epistemic uncertainty quantification in machine learning: The case of direct uncertainty estimation through loss minimisation, Departmental Colloquium, Department of Statistics, LMU, Germany, 2022
  • Preference-based bandit algorithms, DAGStat, Hamburg, Germany, 2022
  • Preference-based bandit algorithms, European Conference on Data Analysis (ECDA), digital conference, 2021
  • Identification of the generalized Condorcet winner in multi-dueling bandits, Advances in Neural Information Processing Systems (NeurIPS), digital conference, 2021
  • Preselection bandits, International Conference on Machine Learning (ICML), digital conference, 2020
  • Confidence sets for change-point problems in nonparametric regression, seminar of the working group for probability theory, Paderborn, Germany, 2019
  • Adaptive confidence sets for kink-location and kink-size in nonparametric regression, DAGStat, Munich, Germany, 2019
  • Adaptive confidence intervals for kink estimation, European Conference on Data Analysis (ECDA), Paderborn, Germany, 2018
  • Construction of asymptotic confidence bands for the jump curve in bivariate regression problems, German Probability and Statistics Days, Freiburg, Germany, 2018
  • Construction of asymptotic confidence bands for the jump curve in bivariate regression problems, Doctoral Students Meeting of the Stochastics Group, Kaiserslautern, Germany, 2017

Posters

  • AC-Band: A combinatorial bandit-based approach to algorithm configuration, AAAI Conference on Artificial Intelligence, Washington, USA, 2023
  • Finding optimal arms in non-stochastic combinatorial bandits with semi-bandit feedback and finite budget, Advances in Neural Information Processing Systems (NeurIPS), New Orleans, USA, 2022
  • Pitfalls of epistemic uncertainty quantification through loss minimisation, Advances in Neural Information Processing Systems (NeurIPS), New Orleans, USA, 2022
  • Preference-based bandit algorithms, MCML Kick-off 2022, Munich, Germany, 2022
  • Machine learning for online algorithm selection under censored feedback, AAAI Conference on Artificial Intelligence (AAAI), digital conference, 2022
  • Preference-based bandit algorithms, All-Hands-Meeting 2022, Munich, Germany, 2022
  • Identification of the generalized Condorcet winner in multi-dueling bandits, Advances in Neural Information Processing Systems (NeurIPS), digital conference, 2021
  • On testing transitivity in online preference learning, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD), digital conference, 2021
  • Single player Monte-Carlo tree search based on the Plackett-Luce model, AAAI Conference on Artificial Intelligence (AAAI), digital conference, 2021
  • Testification of Condorcet winners in dueling bandits, Conference on Uncertainty in Artificial Intelligence (UAI), digital conference, 2021
  • Preselection bandits, Multi Armed Bandit Workshop, London, UK, 2019

Teaching Activities

  • Lecture: Online Machine Learning and Bandits, LMU Munich, WiSe 2022/2023
  • Project Group: Prediction of the KPI Seat Load Factor, LMU Munich, WiSe 2022/2023 (joint project with Lufthansa Group)
  • Lecture: Advanced Machine Learning, LMU Munich, SuSe 2022
  • Lecture: Supervised Learning, LMU Munich, SuSe 2022
  • Tutorial: Uncertainty in Artificial Intelligence and Machine Learning, LMU Munich, WiSe 2021/2022
  • Project Group: Multi armed bandit algorithms, Paderborn University, SuSe and WiSe 2020/2021
  • Lecture: Online and Adaptive Machine Learning, Paderborn University, WiSe 2019/2020 and 2020/2021
  • Lecture: Data Mining, Paderborn University, SuSe 2019 and 2020
  • Tutorials/ Exercise: Praktikum zur Stochastik, University of Marburg, WiSe 2017/2018 and SoSe 2018
  • Tutorials/ Exercise: Quantitatives Risikomanagement, University of Marburg, WiSe 2016/2017
  • Tutorials/ Exercise: Mathematik II (Einführung in die Analysis), University of Marburg, SuSe 2016
  • Tutorials/ Exercise: Nichtparametrische Statistik, University of Marburg, WiSe 2015/2016
  • Tutorials/ Exercise: Einführung in die stochastische Analysis, University of Marburg, SuSe 2015

Mentoring and Theses (Co-)Supervision

  • 4 Ph.D. Candidates, Paderborn University, LMU Munich, 2019 - present
  • 4 Master’s theses, University of Marburg, Paderborn University, LMU Munich, 2016 - 2022
  • 10 Bachelor’s theses, University of Marburg, LMU Munich, 2015 - 2022

Scholarship and Honors