Delayed Reinforcement Learning by Imitation
This addresses a practical issue in reinforcement learning for domains like robotics and trading where delays can cause classic methods to fail, offering a novel solution with theoretical and empirical validation.
The paper tackles the problem of reinforcement learning when agent observations or interactions are delayed, proposing a new algorithm called Delayed Imitation with Dataset Aggregation (DIDA) that uses imitation learning from undelayed demonstrations to act in delayed environments, achieving high performance with remarkable sample efficiency across tasks like robotic locomotion, classic control, and trading.
When the agent's observations or interactions are delayed, classic reinforcement learning tools usually fail. In this paper, we propose a simple yet new and efficient solution to this problem. We assume that, in the undelayed environment, an efficient policy is known or can be easily learned, but the task may suffer from delays in practice and we thus want to take them into account. We present a novel algorithm, Delayed Imitation with Dataset Aggregation (DIDA), which builds upon imitation learning methods to learn how to act in a delayed environment from undelayed demonstrations. We provide a theoretical analysis of the approach that will guide the practical design of DIDA. These results are also of general interest in the delayed reinforcement learning literature by providing bounds on the performance between delayed and undelayed tasks, under smoothness conditions. We show empirically that DIDA obtains high performances with a remarkable sample efficiency on a variety of tasks, including robotic locomotion, classic control, and trading.