Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2501.17115 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908850971475968 |
|---|---|
| author | Hosseinkhan-Boucher, Rémy Semeraro, Onofrio Mathelin, Lionel |
| author_facet | Hosseinkhan-Boucher, Rémy Semeraro, Onofrio Mathelin, Lionel |
| contents | The generalisation and robustness properties of policies learnt through Maximum-Entropy Reinforcement Learning are investigated on chaotic dynamical systems with Gaussian noise on the observable. First, the robustness under noise contamination of the agent's observation of entropy regularised policies is observed. Second, notions of statistical learning theory, such as complexity measures on the learnt model, are borrowed to explain and predict the phenomenon. Results show the existence of a relationship between entropy-regularised policy optimisation and robustness to noise, which can be described by the chosen complexity measures. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_17115 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Evidence on the Regularisation Properties of Maximum-Entropy Reinforcement Learning Hosseinkhan-Boucher, Rémy Semeraro, Onofrio Mathelin, Lionel Machine Learning The generalisation and robustness properties of policies learnt through Maximum-Entropy Reinforcement Learning are investigated on chaotic dynamical systems with Gaussian noise on the observable. First, the robustness under noise contamination of the agent's observation of entropy regularised policies is observed. Second, notions of statistical learning theory, such as complexity measures on the learnt model, are borrowed to explain and predict the phenomenon. Results show the existence of a relationship between entropy-regularised policy optimisation and robustness to noise, which can be described by the chosen complexity measures. |
| title | Evidence on the Regularisation Properties of Maximum-Entropy Reinforcement Learning |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2501.17115 |