In the past years, reinforcement learning (RL) gained much attention in related research communities and public media. It has been shown that it can be used for various types of problems to automate the solution of decision processes. Due to its universality, many other research areas can benefit from the tools provided by RL. My research is focused on the development of RL and related methods for application in revenue management and operations research. Within that area, many recurring decision problems arise. Modern enterprise systems offer the opportunity to access data related to those problems to a former unexperienced extent.

• Evaluation of specific transfer learning mechanisms in the context of pricing problems: This aims at using data from multiple sources to overcome the problem of being unable to generate the amounts of information necessary to run deep reinforcement learning algorithms.
• Multi agent setups with partial or full information: Past research has shown that it is hard to achieve stable equilibria with multiple reinforcement learning based agents in the same system, especially if they rely on non-linear function approximation tools. This project aims at evaluating whether stable equilibria can be achieved when one of the agents has more information available than the others.
• Active state exploration: Past mechanisms that were designed for reinforcement learning algorithms did aim at evaluating if certain past observations might be more important for the learning process than others. During the experiments for this projects, we evaluate if the learning process can be shortened, if we focus on aiming for states the agent currently has only sparse information about.

• Dynamic Programming and Reinforcement Learning Lecture
• Bachelor's Project 2021

• Dynamic Pricing under Competition using Reinforcement Learning, Euro 2021 Athens
• Upcoming talk: Towards Transfer Learning for Revenue and Pricing Management, OR Conference Bern 2021, GOR e.V.
• Talks in cooperation with the HPI research schools, including their weekly presentations and at the University of California, Irvine