Matthias Oberlechner

E-Mail: matthias.oberlechner@tum.de
Phone: +49 (0) 89 289 - 17506

Office: Room 01.10.055
Boltzmannstr. 3
85748 Garching bei München, Germany

Hours: by arrangement


Short Bio

After receiving my M.Sc. in Mathematics at the Technical University of Munich, I started my my Ph.D. in the computer science department under the supervision of Martin Bichler. In my research, I investigate learning dynamics in multi-agent systems with application to economic settings such as auctions and contests. In particular, I am interested in the underlying structures that allow simple learning algorithms to reach the equilibrium strategies in these games. In most of my projects I rely on methods from online convex optimization and numerical simulations.

Education

  • Ph.D. Student in the Department of Computer Science at TUM
    Munich, Germany, since 09/2020
  • M.Sc. Mathematics at TUM
    Munich, Germany, 04/2017 - 06/2020
  • B.Sc. Mathematics at TUM
    Munich, Germany, 10/2013 - 01/2017

Other Experiences

 


Publications

Working Paper

  • On the Convergence of Learning Algorithms in Bayesian Auction Games
    Martin Bichler, Stephan B. Lunowa, Matthias Oberlechner, Fabian R. Pieroth, Barabara Wohlmuth
    Preprint, 2023 [ arXiv ]

Journal Publications

  • Computing Bayes Nash Equilibrium Strategies in Auction Games via Simultaneous Online Dual Averaging
    Martin Bichler, Maximilian Fichtl, Matthias Oberlechner
    Operations Research, 2023 [ DOI | arXiv]
     
  • Learning equilibrium in bilateral bargaining games
    Martin Bichler, Nils Kohring, Matthias Oberlechner, Fabian R. Pieroth
    European Journal of Operational Research, 2023 [ DOI ]

Conference Papers

  • Low Revenue in Display Ad Auctions: Algorithmic Collusion vs. Non-Quasilinear Preferences
    Martin Bichler, Alok Gupta, Laura Mathews, Matthias Oberlechner
    Conference on Information Systems and Technology (CIST), 2024, Seattle (USA). [ arXiv ]
    Presented at International Conference on Operations Research 2024 (OR2024), 2024, Munich (Germany).
    Previous version presented at Ninth Marketplace Innovation Workshop (MIW), 2024, Online.
     
  • Computing Bayes Nash Equilibrium Strategies in Auction Games via Simultaneous Online Dual Averaging
    Martin Bichler, Maximilian Fichtl, Matthias Oberlechner
    24th ACM Conference on Economics and Computation (ACM-EC), 2023, London (UK). [ DOI | arXiv ]

Workshop Papers

  • Computing Bayes Nash Equilibrium Strategies in Crowdsourcing Contests
    Martin Bichler, Markus Ewert, Matthias Oberlechner
    In 32nd Workshop on Information Technologies and Systems (WITS-22), 2022, Copenhagen (Denmark).
     
  • Computing Distributional Bayes Nash Equilibria in Auction Games via Gradient Dynamics
    Martin Bichler, Maximilian Fichtl, Matthias Oberlechner
    In Workshop on Reinforcement Learning in Games (AAAI-22), 2022, Online. [ venue | pdf ]
    Previous version in Workshop on Learning in Presence of Strategic Behavior (NeurIPS21), 2021, Online.

Teaching

Classes

  • Operations Research (IN0024)
    Teaching Assistant: SS21, SS22, SS23, SS24
     
  • Auction Theory and Market Design (IN2211)
    Teaching Assistant: WS21/22, WS22/23

Supervised Theses

  • Learning Bayes-Nash Equilibrium Strategies in Auctions using Discontinuous Neural Networks
    Thesis, Bachelor - Information Systems, 2024
  • Analyzing the Potentialness of Random Games using the Hodge Decomposition
    Thesis, Bachelor - Informatics, 2024
  • Numerical Methods for Variational Inequalities and Application to Equilibrium Learning in Game Theory
    Thesis, Master - Mathematics in Data Science, 2024
  • Learning in Continuous Games with Bandit Feedback
    Thesis, Bachelor - Information Systems, 2024
  • Numerical Experiments for Equilibrium Learning with Bandit-Feedback in Contests
    Thesis, Master - Informatics, 2023
  • Equilibrium Learning in Bertrand and Cournot Competition
    Interdisciplinary Project (IDP), Master - Informatics, 2023
  • Development of a DGS Auction Problem Generator
    Thesis, Bachelor - Information Systems, 2023
  • Application-oriented Analysis of Inventory Management in the Context of Strategic Portfolio Planning in the Automotive Sector Using Operations Research
    Thesis, Bachelor - Information Systems, 2023
  • Learning Discrete Equilibrium Strategies in Auction Games
    Thesis, Master - Robotics, Cognition, Intelligence, 2022
  • Visualisation of different Learning Algorithms in Matrix Games
    Thesis, Bachelor - Information Systems, 2022
  • Multiplicative Weights Update in Congestion Games
    Thesis, Bachelor Information Systems, 2022
  • Solving the CVRPTW: Quantum Annealing vs. Exact Optimization
    Thesis, Master - Mathematics, 2022
  • Visualization of Different Equilibrium Concepts in Matrix Games
    Thesis, Bachelor - Information Systems, 2022
  • A Comparison of No-External-Regret and No-Internal-Regret Learning Algorithmis in Matrix Games
    Thesis, Bachelor - Informatics, 2022
  • No-Regret Learning in Finite Games
    Thesis, Bachelor - Information Systems, 2022
  • Stable Marriage Problem: Fair Algorithms and Applications
    Thesis, Bachelor - Information Systems, 2021

For available theses topics, check out this page.