| layout | author_profile |
|---|---|
splash |
true |
We are a group of researches from the universities of Freiburg and Hannover working in this domain and we want this page to be a resource for not just our own research but the state of AutoRL as a whole. We write an AutoRL blog and try to keep on top of interesting events for AutoRL enthusiasts.
Reinforcement learning (RL) is a simple, yet powerful paradigm for training intelligent agents to perform a given task. To do so, RL agents interact with the world they exist in. Guided by a reward signal, RL agents learn in a trial-and-error fashion. That is, RL agents follow a policy, observe if the policy was good or bad and, depending on this outcome, update their policy to get better at a given task.
The simplicity of this paradigm raises the expectations that RL should be applicable in a wide variety of problem domains. However, in practice, it is well known that existing RL algorithms are brittle, require attention to minute implementation details and are sensitive to the experimental setup in general. As a result, RL has not found wide-spread adoption for many real-world tasks. AutoML provides ample solution approaches to overcome these issues in RL, but conversely, RL also offers novel opportunities for AutoML researchers.
- Theresa Eimer and André Biedenkapp will be giving a series of tutorials on AutoRL in September and October:
- 06.09.24 AutoML School
- 10.09.24 AutoML'24 Conference
- 18.10.24 ECAI'24
- We are excited to announce the first AutoRL workshop at ICML 2024, Vienna. Please visit the workshop website for more details.