☆ 4.7 Article

Achieving efficient interpretability of reinforcement learning via policy distillation and selective input gradient regularization

NEURAL NETWORKS (2023)

The intelligent critic framework for advanced optimal control

Ding Wang et al.

Summary: The idea of optimization is important in many disciplines, especially for advanced control design based on artificial intelligence. Due to the difficulty of solving optimal control problems for nonlinear systems, new learning strategies with intelligent components are necessary. The rapid development of computer and networked techniques has also promoted research on optimal control in the discrete-time domain. This paper presents the bases, derivation, and recent progresses of critic intelligence for discrete-time optimal control design, with a focus on the iterative framework.

ARTIFICIAL INTELLIGENCE REVIEW (2022)

添加到收藏夹

Article Multidisciplinary Sciences