SBEED: Convergent Reinforcement Learning With Nonlinear Function Approximation

Abstract

When function approximation is used, solving the Bellman optimality equation with stability guarantees has remained a major open problem in reinforcement learning for decades. The fundamental difficulty is that the Bellman operator may become an expansion in general, resulting in oscillating and even divergent behavior of popular algorithms like Q-learning. In this paper, we revisit the Bellman equation, and reformulate it into a novel primal-dual optimization problem using Nesterov's smoothing technique and the Legendre-Fenchel transformation. We then develop a new algorithm, called Smoothed Bellman Error Embedding, to solve this optimization problem where any differentiable function class may be used. We provide what we believe to be the first convergence guarantee for general nonlinear function approximation, and analyze the algorithm's sample complexity. Empirically, our algorithm compares favorably to state-of-the-art baselines in several benchmark control problems.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keydai2017sbeed

Related papers

• Non-stationary Reinforcement Learning Under General Function Approximation (2023)0.00
• Provably Efficient Reinforcement Learning With Linear Function Approximation (2019)11.76
• Distributional Reinforcement Learning With Linear Function Approximation (2019)0.00
• Uniform-pac Bounds For Reinforcement Learning With Linear Function Approximation (2021)0.00
• Improved Regret For Efficient Online Reinforcement Learning With Linear Function Approximation (2023)0.00
• Reinforcement Learning With Unbiased Policy Evaluation And Linear Function Approximation (2022)0.00
• Continuous-time Reinforcement Learning: Ellipticity Enables Model-free Value Function Approximation (2026)0.00
• Stabilizing Q-learning With Linear Architectures For Provably Efficient Learning (2022)0.00