Quasi-Newton Iteration in Deterministic Policy Gradient.
Arash Bahari KordabadHossein Nejatbakhsh EsfahaniWenqi CaiSébastien GrosPublished in: ACC (2022)
Keyphrases
- quasi newton
- policy gradient
- gradient method
- step size
- optimization methods
- optimization method
- reinforcement learning
- function approximation
- newton method
- convergence rate
- partially observable markov decision processes
- approximation methods
- genetic algorithm
- reinforcement learning algorithms
- least squares
- markov decision processes
- state space
- recommender systems
- feature space