Anytime Incremental $ρ$POMDP Planning in Continuous Spaces
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916597472428032 |
|---|---|
| author | Benchetrit, Ron Lev-Yehudi, Idan Zhitnikov, Andrey Indelman, Vadim |
| author_facet | Benchetrit, Ron Lev-Yehudi, Idan Zhitnikov, Andrey Indelman, Vadim |
| contents | Partially Observable Markov Decision Processes (POMDPs) provide a robust framework for decision-making under uncertainty in applications such as autonomous driving and robotic exploration. Their extension, $ρ$POMDPs, introduces belief-dependent rewards, enabling explicit reasoning about uncertainty. Existing online $ρ$POMDP solvers for continuous spaces rely on fixed belief representations, limiting adaptability and refinement - critical for tasks such as information-gathering. We present $ρ$POMCPOW, an anytime solver that dynamically refines belief representations, with formal guarantees of improvement over time. To mitigate the high computational cost of updating belief-dependent rewards, we propose a novel incremental computation approach. We demonstrate its effectiveness for common entropy estimators, reducing computational cost by orders of magnitude. Experimental results show that $ρ$POMCPOW outperforms state-of-the-art solvers in both efficiency and solution quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_02549 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Anytime Incremental $ρ$POMDP Planning in Continuous Spaces Benchetrit, Ron Lev-Yehudi, Idan Zhitnikov, Andrey Indelman, Vadim Artificial Intelligence Machine Learning Robotics Partially Observable Markov Decision Processes (POMDPs) provide a robust framework for decision-making under uncertainty in applications such as autonomous driving and robotic exploration. Their extension, $ρ$POMDPs, introduces belief-dependent rewards, enabling explicit reasoning about uncertainty. Existing online $ρ$POMDP solvers for continuous spaces rely on fixed belief representations, limiting adaptability and refinement - critical for tasks such as information-gathering. We present $ρ$POMCPOW, an anytime solver that dynamically refines belief representations, with formal guarantees of improvement over time. To mitigate the high computational cost of updating belief-dependent rewards, we propose a novel incremental computation approach. We demonstrate its effectiveness for common entropy estimators, reducing computational cost by orders of magnitude. Experimental results show that $ρ$POMCPOW outperforms state-of-the-art solvers in both efficiency and solution quality. |
| title | Anytime Incremental $ρ$POMDP Planning in Continuous Spaces |
| topic | Artificial Intelligence Machine Learning Robotics |
| url | https://arxiv.org/abs/2502.02549 |