- Title: Planning and learning
- Speaker: Piotr Miłoś (Polish Academy of Sciences)
- Time and date: 1pm to 2pm, September 15th, 2021 (Wednesday)
- Room: Virtual (Zoom)
The Game AI Research Group is glad to announce a (virtual) talk by Piotr Miłoś on Wednesday September 15th at 13:00.
All welcome (especially students), no pre-booking required
Abstract
Planning and search have been immensely studied topics in AI research. However, still many questions stay unanswered. In my talk, I will explore topics on the intersection of planning and learning. A general question is how learning methods can be used to make planning more efficient. A closely related one is how to develop planning methods robust to (unavoidable) errors of learned models.
I will present Subgoal Search and Entropy Search Tree based on our recent work [1] [2].
- [1] Subgoal Search For Complex Reasoning Tasks, arXiv2108.11204
- [2] Robust and Efficient Planning using Adaptive Entropy Tree Search, arXiv2102.06808
Bio
Piotr Milos is an associate professor at the Polish Academy of Sciences, where he leads the Awarelab research group (https://awarelab.gitlab.io/website/). The group focuses on machine learning topics with an emphasis on reinforcement learning. Before working in ML prof. Milos specialized in probability theory.