Non-Deterministic Policy Improvement Stabilizes Approximated Reinforcement Learning
This addresses stability issues in reinforcement learning algorithms, potentially benefiting practitioners using approximated methods, though it appears incremental.
The paper tackles instability in approximated reinforcement learning caused by greedy policy improvement, showing that non-deterministic policy improvement stabilizes methods like LSPI, with a suitable value function representation providing additional stabilization.
This paper investigates a type of instability that is linked to the greedy policy improvement in approximated reinforcement learning. We show empirically that non-deterministic policy improvement can stabilize methods like LSPI by controlling the improvements' stochasticity. Additionally we show that a suitable representation of the value function also stabilizes the solution to some degree. The presented approach is simple and should also be easily transferable to more sophisticated algorithms like deep reinforcement learning.