Doctoral Students at the School of Business Informatics and Mathematics

Topic/title: Optimization on Random Functions as a Model for Optimization in Machine Learning
Summary: Machine learning looks for the model that best fits the existing data. To better understand this high-dimensional optimization problem, I am studying the behavior of optimization algorithms on random functions.

Topic/title: Cartel Law Compliant AI
Summary: Nowadays, online stores often use AIs to set their prices. Could these AIs form cartels? And how could we detect and prevent it?

Topic/title: Equilibrium Seeking through the Lens of Dynamical Systems Theory
Summary: We aim to advance computational techniques for hierarchal optimization problems and games. Our methods will be applied to energy systems and computational imaging.

Topic/title: Evaluating AI-Based Policies with Social Simulation
Summary: AI is increasingly being used in public spaces (for instance, as part of smart city schemes). I am using simulations to study the (possibly unintended) effects of this development.

Topic/title: Stochastic Processes and Reinforcement Learning
Summary: Reinforcement learning (RL) algorithms are used to train AIs like ChatGPT. My dissertation investigates the convergence of RL algorithms to establish theoretical guarantees for their application.

Topic/title: Methods of Causal Discovery in AI
Summary: Why does an AI predict X instead of Y for Alice? What would you need to change for it to predict Y? I am developing methods to answer these questions.

Researched and edited by: Rheia Martiny/December 2023