Representation of Reinforcement Learning Policies in Reproducing Kernel Hilbert Spaces
This provides a stable policy representation framework for reinforcement learning tasks, particularly beneficial in applications requiring robustness, though it appears incremental as it builds on existing RKHS methods.
The authors tackled the problem of representing reinforcement learning policies with theoretical guarantees by embedding them in a low-dimensional reproducing kernel Hilbert space (RKHS), achieving almost no decrease in return in experiments on classic RL domains.
We propose a general framework for policy representation for reinforcement learning tasks. This framework involves finding a low-dimensional embedding of the policy on a reproducing kernel Hilbert space (RKHS). The usage of RKHS based methods allows us to derive strong theoretical guarantees on the expected return of the reconstructed policy. Such guarantees are typically lacking in black-box models, but are very desirable in tasks requiring stability. We conduct several experiments on classic RL domains. The results confirm that the policies can be robustly embedded in a low-dimensional space while the embedded policy incurs almost no decrease in return.