Closing the Gap: Achieving Global Convergence (Last Iterate) of Actor-Critic under Markovian Sampling with Neural Network Parametrization.
Mudit GaurAmrit Singh BediDi WangVaneet AggarwalPublished in: CoRR (2024)
Keyphrases
- global convergence
- neural network
- actor critic
- convergence rate
- gradient method
- optimization methods
- global optimum
- convergence speed
- reinforcement learning
- temporal difference
- policy gradient
- approximate dynamic programming
- optimal control
- step size
- back propagation
- artificial neural networks
- neuro fuzzy
- monte carlo
- policy iteration
- hybrid algorithm
- particle swarm
- reinforcement learning algorithms
- genetic algorithm
- sample size
- simulated annealing
- recurrent neural networks
- fuzzy logic
- function approximation
- particle swarm optimization algorithm
- particle swarm optimization
- optimization problems