Hill Climbing on Value Estimates for Search-control in Dyna
This work addresses a key bottleneck in Dyna for reinforcement learning practitioners, offering an incremental improvement in search-control mechanisms.
The paper tackles the problem of search-control in Dyna, a model-based reinforcement learning architecture, by proposing to generate states via hill climbing on the current value function estimate, which propagates value from high-value regions and updates likely future states. The result is HC-Dyna, which demonstrates significant sample efficiency improvements in four classical domains.
Dyna is an architecture for model-based reinforcement learning (RL), where simulated experience from a model is used to update policies or value functions. A key component of Dyna is search-control, the mechanism to generate the state and action from which the agent queries the model, which remains largely unexplored. In this work, we propose to generate such states by using the trajectory obtained from Hill Climbing (HC) the current estimate of the value function. This has the effect of propagating value from high-value regions and of preemptively updating value estimates of the regions that the agent is likely to visit next. We derive a noisy projected natural gradient algorithm for hill climbing, and highlight a connection to Langevin dynamics. We provide an empirical demonstration on four classical domains that our algorithm, HC-Dyna, can obtain significant sample efficiency improvements. We study the properties of different sampling distributions for search-control, and find that there appears to be a benefit specifically from using the samples generated by climbing on current value estimates from low-value to high-value region.