Stochastic Approximation With Iterate-Dependent Markov Noise Under Verifiable Conditions in Compact State Space With the Stability of Iterates Not Ensured.
Prasenjit KarmakarShalabh BhatnagarPublished in: IEEE Trans. Autom. Control. (2021)
Keyphrases
- stochastic approximation
- state space
- markov chain
- sufficient conditions
- monte carlo
- reinforcement learning
- markov decision processes
- policy iteration
- state variables
- optimal policy
- dynamical systems
- dynamic programming
- finite state
- action space
- temporal difference learning
- multi agent
- particle filter
- neural network
- reinforcement learning algorithms
- markov decision process
- machine learning