Vortrag
Reinforcement Learning for Robotics
- von -
- Halle C5C5.100
- Sprache: Englisch
- Vortragstyp: Workshop
Vortragsbeschreibung
Hands-on workshop training an AI agent from scratch to balance a small wheeled robot using the MuJoCo physics simulator, Python and the PPO algorithm, covering domain randomization, curriculum learning and the sim-to-real transfer problem, then running the trained policy on real hardware.
