Global convergence of natural policy gradient with Hessian-aided momentum variance reduction

DOI10.1007/s10915-024-02688-xMaRDI QIDQ6629222

Jinchi Chen, Jie Feng, Ke Wei

Publication date: 29 October 2024

Published in: Journal of Scientific Computing (Search for Journal in Brave)

zbMATH Keywords

variance reductionreinforcement learningsample complexitynatural policy gradient

Mathematics Subject Classification ID

Nonconvex programming, global optimization (90C26) Methods of quasi-Newton type (90C53) Stochastic learning and adaptive control (93E35) Stochastic systems in control theory (general) (93E03) Computational methods for problems pertaining to operations research and mathematical programming (90-08)

Cites Work

• Unnamed Item
• Unnamed Item
• Policy mirror descent for reinforcement learning: linear convergence, new sampling complexity, and generalized problem classes
• OnActor-Critic Algorithms
• Fast Global Convergence of Natural Policy Gradient Methods with Entropy Regularization
• Global Convergence of Policy Gradient Methods to (Almost) Locally Optimal Policies
• Smoothing policies and safe policy gradients
• Policy Mirror Descent for Regularized Reinforcement Learning: A Generalized Framework with Linear Convergence
• Homotopic policy mirror descent: policy convergence, algorithmic regularization, and improved sample complexity

This page was built for publication: Global convergence of natural policy gradient with Hessian-aided momentum variance reduction