Efficient Entropy for Policy Gradient with Multidimensional Action Space
This addresses a computational bottleneck for researchers and practitioners in reinforcement learning dealing with high-dimensional action spaces, though it is incremental as it builds on existing entropy bonus methods.
The paper tackled the computational infeasibility of calculating entropy and its gradient for exploration in policy gradient methods with high-dimensional discrete action spaces by developing novel unbiased estimators, resulting in substantial performance improvements with marginal additional computational cost in tested environments like a multi-hunter multi-rabbit grid game and a multi-agent multi-arm bandit problem.
In recent years, deep reinforcement learning has been shown to be adept at solving sequential decision processes with high-dimensional state spaces such as in the Atari games. Many reinforcement learning problems, however, involve high-dimensional discrete action spaces as well as high-dimensional state spaces. This paper considers entropy bonus, which is used to encourage exploration in policy gradient. In the case of high-dimensional action spaces, calculating the entropy and its gradient requires enumerating all the actions in the action space and running forward and backpropagation for each action, which may be computationally infeasible. We develop several novel unbiased estimators for the entropy bonus and its gradient. We apply these estimators to several models for the parameterized policies, including Independent Sampling, CommNet, Autoregressive with Modified MDP, and Autoregressive with LSTM. Finally, we test our algorithms on two environments: a multi-hunter multi-rabbit grid game and a multi-agent multi-arm bandit problem. The results show that our entropy estimators substantially improve performance with marginal additional computational cost.