StochMod EURO PhD School
Reinforcement Learning Applied to Operations Research
17–24 July 2022
This PhD school introduced students in stochastic modelling to a range of reinforcement learning (RL) techniques. Participants applied these methods to solve operations research problems in a hackathon-style workshop.
The program included lectures from leading specialists, delivered both on-site and online, as well as dedicated time for collaborative work in small groups.
Participants were divided into groups based on their preferred application areas, including inventory management, queueing systems, and pricing. Following an introduction to Markov decision processes, they identified suitable problems and tackled them using techniques such as temporal-difference learning, deep reinforcement learning, and multi-armed bandits.
Through guided tutorials, students implemented and compared different algorithms. The resulting analyses and findings were then presented and discussed with experts.
The facilities at Castle Gimborn provided an excellent environment for focused work, productive discussions, and informal networking alongside the official program. The social activities included a walking tour and a dinner in a traditional German beer garden.
We look forward to seeing participants present work inspired by the school at future conferences.