RUDDER: Return Decomposition for Delayed Rewards
Jose A. Arjona-Medina, Michael Gillhofer, Michael Widrich, Thomas Unterthiner, Johannes Brandstetter, Sepp Hochreiter
–Neural Information Processing Systems
We propose RUDDER, a novel reinforcement learning approach for delayed rewards in finite Markov decision processes (MDPs).
Neural Information Processing Systems
Oct-2-2025, 05:30:57 GMT
- Country:
- North America > United States (0.46)
- Europe (0.46)
- Industry:
- Education (0.46)
- Leisure & Entertainment > Games (0.30)
- Technology: