Contextual Pre-planning on Reward Machine Abstractions for Enhanced Transfer in Deep Reinforcement Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909113849479168 |
|---|---|
| author | Azran, Guy Danesh, Mohamad H. Albrecht, Stefano V. Keren, Sarah |
| author_facet | Azran, Guy Danesh, Mohamad H. Albrecht, Stefano V. Keren, Sarah |
| contents | Recent studies show that deep reinforcement learning (DRL) agents tend to overfit to the task on which they were trained and fail to adapt to minor environment changes. To expedite learning when transferring to unseen tasks, we propose a novel approach to representing the current task using reward machines (RMs), state machine abstractions that induce subtasks based on the current task's rewards and dynamics. Our method provides agents with symbolic representations of optimal transitions from their current abstract state and rewards them for achieving these transitions. These representations are shared across tasks, allowing agents to exploit knowledge of previously encountered symbols and transitions, thus enhancing transfer. Empirical results show that our representations improve sample efficiency and few-shot transfer in a variety of domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2307_05209 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Contextual Pre-planning on Reward Machine Abstractions for Enhanced Transfer in Deep Reinforcement Learning Azran, Guy Danesh, Mohamad H. Albrecht, Stefano V. Keren, Sarah Artificial Intelligence Machine Learning Recent studies show that deep reinforcement learning (DRL) agents tend to overfit to the task on which they were trained and fail to adapt to minor environment changes. To expedite learning when transferring to unseen tasks, we propose a novel approach to representing the current task using reward machines (RMs), state machine abstractions that induce subtasks based on the current task's rewards and dynamics. Our method provides agents with symbolic representations of optimal transitions from their current abstract state and rewards them for achieving these transitions. These representations are shared across tasks, allowing agents to exploit knowledge of previously encountered symbols and transitions, thus enhancing transfer. Empirical results show that our representations improve sample efficiency and few-shot transfer in a variety of domains. |
| title | Contextual Pre-planning on Reward Machine Abstractions for Enhanced Transfer in Deep Reinforcement Learning |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2307.05209 |