APRIL: Active Preference-learning based Reinforcement Learning
This addresses the challenge of RL in domains like swarm robotics where human experts can only provide preferences, but it is incremental as it builds on existing preference-based RL frameworks.
The paper tackles the problem of reinforcement learning when expert-designed rewards or demonstrations are unavailable, by combining preference-based RL with active ranking to reduce the number of expert queries needed. Experiments show that a couple of dozen rankings can learn a competent policy in testbeds like mountain car and cancer treatment.
This paper focuses on reinforcement learning (RL) with limited prior knowledge. In the domain of swarm robotics for instance, the expert can hardly design a reward function or demonstrate the target behavior, forbidding the use of both standard RL and inverse reinforcement learning. Although with a limited expertise, the human expert is still often able to emit preferences and rank the agent demonstrations. Earlier work has presented an iterative preference-based RL framework: expert preferences are exploited to learn an approximate policy return, thus enabling the agent to achieve direct policy search. Iteratively, the agent selects a new candidate policy and demonstrates it; the expert ranks the new demonstration comparatively to the previous best one; the expert's ranking feedback enables the agent to refine the approximate policy return, and the process is iterated. In this paper, preference-based reinforcement learning is combined with active ranking in order to decrease the number of ranking queries to the expert needed to yield a satisfactory policy. Experiments on the mountain car and the cancer treatment testbeds witness that a couple of dozen rankings enable to learn a competent policy.