Rank-1 Approximation of Inverse Fisher for Natural Policy Gradients in Deep Reinforcement Learning
This addresses efficiency for deep reinforcement learning practitioners, but it is incremental as it builds on existing natural gradient methods.
The paper tackles the computational bottleneck of inverting the Fisher Information Matrix in natural policy gradients by proposing a rank-1 approximation method, achieving superior performance over actor-critic and trust-region baselines in diverse environments.
Natural gradients have long been studied in deep reinforcement learning due to their fast convergence properties and covariant weight updates. However, computing natural gradients requires inversion of the Fisher Information Matrix (FIM) at each iteration, which is computationally prohibitive in nature. In this paper, we present an efficient and scalable natural policy optimization technique that leverages a rank-1 approximation to full inverse-FIM. We theoretically show that under certain conditions, a rank-1 approximation to inverse-FIM converges faster than policy gradients and, under some conditions, enjoys the same sample complexity as stochastic policy gradient methods. We benchmark our method on a diverse set of environments and show that it achieves superior performance to standard actor-critic and trust-region baselines.