Zap Q-Learning With Nonlinear Function Approximation
This addresses the challenge of slow convergence in reinforcement learning for practitioners, though it is incremental as it builds on existing Zap Q-learning methods.
The paper tackled the problem of accelerating convergence in Zap Q-learning by extending stability theory beyond restrictive settings, showing consistency with nonlinear function approximation and achieving quick convergence in tests on OpenAI Gym examples.
Zap Q-learning is a recent class of reinforcement learning algorithms, motivated primarily as a means to accelerate convergence. Stability theory has been absent outside of two restrictive classes: the tabular setting, and optimal stopping. This paper introduces a new framework for analysis of a more general class of recursive algorithms known as stochastic approximation. Based on this general theory, it is shown that Zap Q-learning is consistent under a non-degeneracy assumption, even when the function approximation architecture is nonlinear. Zap Q-learning with neural network function approximation emerges as a special case, and is tested on examples from OpenAI Gym. Based on multiple experiments with a range of neural network sizes, it is found that the new algorithms converge quickly and are robust to choice of function approximation architecture.