Controlled interacting particle algorithms for simulation-based reinforcement learning.
Anant A. JoshiAmirhossein TaghvaeiPrashant G. MehtaSean P. MeynPublished in: Syst. Control. Lett. (2022)
Keyphrases
- reinforcement learning
- learning algorithm
- computational cost
- computational efficiency
- worst case
- times faster
- neural network
- convergence rate
- recently developed
- orders of magnitude
- benchmark datasets
- machine learning algorithms
- computationally efficient
- optimization problems
- multi agent
- data sets
- model free
- policy search