Don't Forget Your Teacher: A Corrective Reinforcement Learning Framework
This addresses the adoption barrier for RL in real-world settings where small adjustments to existing policies are needed, though it is incremental as it builds on standard RL with added constraints.
The paper tackles the problem of practitioners being reluctant to adopt RL solutions that deviate too much from existing policies, by proposing a student-teacher RL framework that constrains the RL policy to stay close to a teacher policy, such as the status quo, while maximizing reward. Numerical experiments in GridWorld scenarios confirm the method's effectiveness, with the policy proven to be asymptotically optimal.
Although reinforcement learning (RL) can provide reliable solutions in many settings, practitioners are often wary of the discrepancies between the RL solution and their status quo procedures. Therefore, they may be reluctant to adapt to the novel way of executing tasks proposed by RL. On the other hand, many real-world problems require relatively small adjustments from the status quo policies to achieve improved performance. Therefore, we propose a student-teacher RL mechanism in which the RL (the "student") learns to maximize its reward, subject to a constraint that bounds the difference between the RL policy and the "teacher" policy. The teacher can be another RL policy (e.g., trained under a slightly different setting), the status quo policy, or any other exogenous policy. We formulate this problem using a stochastic optimization model and solve it using a primal-dual policy gradient algorithm. We prove that the policy is asymptotically optimal. However, a naive implementation suffers from high variance and convergence to a stochastic optimal policy. With a few practical adjustments to address these issues, our numerical experiments confirm the effectiveness of our proposed method in multiple GridWorld scenarios.