Model-Based Deep RL

Trajectory Optimization

Direct optimization of state-action sequences using the model's gradient to find optimal trajectories, particularly effective for continuous systems.

← Wstecz