Gradient-Aware Model-based Policy Search
This addresses a specific bottleneck in model-based RL for agents, though it appears incremental as it builds on existing methods with a novel weighting scheme.
The paper tackles the problem of poor model estimates in model-based reinforcement learning due to ignoring policy relevance, by introducing a gradient-aware approach that focuses learning on environment portions critical for policy improvement, and empirically validates it on benchmark domains.
Traditional model-based reinforcement learning approaches learn a model of the environment dynamics without explicitly considering how it will be used by the agent. In the presence of misspecified model classes, this can lead to poor estimates, as some relevant available information is ignored. In this paper, we introduce a novel model-based policy search approach that exploits the knowledge of the current agent policy to learn an approximate transition model, focusing on the portions of the environment that are most relevant for policy improvement. We leverage a weighting scheme, derived from the minimization of the error on the model-based policy gradient estimator, in order to define a suitable objective function that is optimized for learning the approximate transition model. Then, we integrate this procedure into a batch policy improvement algorithm, named Gradient-Aware Model-based Policy Search (GAMPS), which iteratively learns a transition model and uses it, together with the collected trajectories, to compute the new policy parameters. Finally, we empirically validate GAMPS on benchmark domains analyzing and discussing its properties.