1 paper · 1 filter
Daniel Palenicek, Michael Lutter, Joao Carvalho +1
Model-based reinforcement learning is one approach to increase sample efficiency. However, the accuracy of the dynamics model and the resulting compounding error over modelled traj…