Sparse Tree Search Optimality Guarantees in POMDPs with Continuous Observation Spaces

@inproceedings{Lim2020SparseTS,
  title={Sparse Tree Search Optimality Guarantees in POMDPs with Continuous Observation Spaces},
  author={M. H. Lim and C. Tomlin and Zachary Sunberg},
  booktitle={IJCAI},
  year={2020}
}
  • M. H. Lim, C. Tomlin, Zachary Sunberg
  • Published in IJCAI 2020
  • Computer Science, Engineering, Mathematics
  • Partially observable Markov decision processes (POMDPs) with continuous state and observation spaces have powerful flexibility for representing real-world decision and control problems but are notoriously difficult to solve. Recent online sampling-based algorithms that use observation likelihood weighting have shown unprecedented effectiveness in domains with continuous observation spaces. However there has been no formal theoretical justification for this technique. This work offers such a… CONTINUE READING

    References

    SHOWING 1-10 OF 23 REFERENCES
    Solving POMDPs with Continuous or Large Discrete Observation Spaces
    • 117
    • PDF
    Planning and Acting in Partially Observable Stochastic Domains
    • 3,582
    • PDF
    What makes some POMDP problems easy to approximate?
    • 80
    • PDF
    Information Space Receding Horizon Control
    • 21
    Belief space planning assuming maximum likelihood observations
    • 298
    • PDF
    Motion planning under uncertainty using iterative local optimization in belief space
    • 239
    • PDF
    DESPOT: Online POMDP Planning with Regularization
    • 271
    • PDF
    The Complexity of Markov Decision Processes
    • 1,242
    • PDF