File(s) under permanent embargo
Weak human preference supervision for deep reinforcement learning
journal contribution
posted on 2023-05-20, 23:38 authored by Cao, Z, Wong, KC, Lin, C-TThe current reward learning from human preferences could be used to resolve complex reinforcement learning (RL) tasks without access to a reward function by defining a single fixed preference between pairs of trajectory segments. However, the judgment of preferences between trajectories is not dynamic and still requires human input over thousands of iterations. In this study, we proposed a weak human preference supervision framework, for which we developed a human preference scaling model that naturally reflects the human perception of the degree of weak choices between trajectories and established a human-demonstration estimator through supervised learning to generate the predicted preferences for reducing the number of human inputs. The proposed weak human preference supervision framework can effectively solve complex RL tasks and achieve higher cumulative rewards in simulated robot locomotion-MuJoCo games-relative to the single fixed human preferences. Furthermore, our established human-demonstration estimator requires human feedback only for less than 0.01% of the agent's interactions with the environment and significantly reduces the cost of human inputs by up to 30% compared with the existing approaches. To present the flexibility of our approach, we released a video (https://youtu.be/jQPe1OILT0M) showing comparisons of the behaviors of agents trained on different types of human input. We believe that our naturally inspired human preferences with weakly supervised learning are beneficial for precise reward learning and can be applied to state-of-the-art RL systems, such as human-autonomy teaming systems.
History
Publication title
IEEE Transactions on Neural Networks and Learning SystemsVolume
32Issue
12Pagination
5369-5378ISSN
2162-237XDepartment/School
School of Information and Communication TechnologyPublisher
Institute of Electrical and Electronics EngineersPlace of publication
United StatesRights statement
Copyright 2021 IEEERepository Status
- Restricted