Acting upon Imagination: when to trust imagined trajectories in model based reinforcement learning

arXiv:2105.05716v5 Announce Type: replace-cross
Abstract: Model-based reinforcement learning (MBRL) aims to learn model(s) of the environment dynamics that can predict the outcome of its actions. Forward application of the model yields so called imagined trajectories (sequences of action, predicted state-reward) used to optimize the set of candidate actions that maximize expected reward. The outcome, an ideal imagined trajectory or plan, is imperfect and typically MBRL relies on model predictive control (MPC) to overcome this by continuously re-planning from scratch, incurring …

abstract acting application arxiv cs.ai cs.lg dynamics environment imagination learn reinforcement reinforcement learning set state the environment trust type