Sparsely Changing Latent States for Prediction and Planning in Partially Observable Domains.
Christian GumbschMartin V. ButzGeorg MartiusPublished in: NeurIPS (2021)
Keyphrases
- partially observable domains
- sensing actions
- partially observable
- planning problems
- situation calculus
- belief state
- event calculus
- partially observable markov decision processes
- ai planning
- reinforcement learning
- decision problems
- domain independent
- initial state
- integer programming
- heuristic search
- random walk
- inverse reinforcement learning
- infinite horizon
- classical planning
- domain specific
- spatio temporal
- search algorithm