Dynamic probabilistic logic models for effective abstractions in RL
This addresses sample efficiency and generalization issues in reinforcement learning for complex environments, though it appears incremental as it builds on prior work (RePReL).
The paper tackles the problem of sample-efficient learning and task transfer in complex reinforcement learning environments by proposing RePReL, a hierarchical framework that uses dynamic probabilistic logic models for state abstraction. The result shows that RePReL achieves better performance and efficient learning on the task at hand, with better generalization to unseen tasks.
State abstraction enables sample-efficient learning and better task transfer in complex reinforcement learning environments. Recently, we proposed RePReL (Kokel et al. 2021), a hierarchical framework that leverages a relational planner to provide useful state abstractions for learning. We present a brief overview of this framework and the use of a dynamic probabilistic logic model to design these state abstractions. Our experiments show that RePReL not only achieves better performance and efficient learning on the task at hand but also demonstrates better generalization to unseen tasks.