Knowledge Transfer through Value Function for Compositional Tasks
Primary Area: reinforcement learning
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: Transfer Learnig, Curriculum Learning, Deep Reinforcement Learning
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
Abstract: Deep Reinforcement Learning methods are sample inefficient when exploring the environment from scratch. In this work, we introduce an approach of knowledge transfer using the value function combined with curriculum learning, which aims to leverage the learning process by transferring knowledge among progressively increasing task complexity. Our main contribution is demonstrating the effectiveness of this approach by modifying the degrees of freedom of the target task, breaking it down into simpler sub-tasks, and transferring the knowledge along the curriculum steps. We empirically demonstrate the broad possibilities of modifying the degrees of freedom of the target task to leverage learning in classical Reinforcement Learning problems and a real-world control task.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 1210
Loading