Generation of Geodesics with Actor-Critic Reinforcement Learning to Predict Midpoints
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911351123738624 |
|---|---|
| author | Kasaura, Kazumi |
| author_facet | Kasaura, Kazumi |
| contents | To find the shortest paths for all pairs on manifolds with infinitesimally defined metrics, we introduce a framework to generate them by predicting midpoints recursively. To learn midpoint prediction, we propose an actor-critic approach. We prove the soundness of our approach and show experimentally that the proposed method outperforms existing methods on several planning tasks, including path planning for agents with complex kinematics and motion planning for multi-degree-of-freedom robot arms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_01991 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Generation of Geodesics with Actor-Critic Reinforcement Learning to Predict Midpoints Kasaura, Kazumi Machine Learning Artificial Intelligence 51F99 (Primary) 53C60, 68T05 (Secondary) I.2.6 To find the shortest paths for all pairs on manifolds with infinitesimally defined metrics, we introduce a framework to generate them by predicting midpoints recursively. To learn midpoint prediction, we propose an actor-critic approach. We prove the soundness of our approach and show experimentally that the proposed method outperforms existing methods on several planning tasks, including path planning for agents with complex kinematics and motion planning for multi-degree-of-freedom robot arms. |
| title | Generation of Geodesics with Actor-Critic Reinforcement Learning to Predict Midpoints |
| topic | Machine Learning Artificial Intelligence 51F99 (Primary) 53C60, 68T05 (Secondary) I.2.6 |
| url | https://arxiv.org/abs/2407.01991 |