Generalized Second-Order Value Iteration in Markov Decision Processes.
Chandramouli KamanchiRaghuram Bharadwaj DiddigiShalabh BhatnagarPublished in: IEEE Trans. Autom. Control. (2022)
Keyphrases
- markov decision processes
- optimal policy
- policy iteration
- finite state
- state space
- transition matrices
- reinforcement learning
- partially observable
- reinforcement learning algorithms
- finite horizon
- dynamic programming
- markov decision process
- decision theoretic planning
- infinite horizon
- risk sensitive
- decision processes
- stochastic shortest path
- planning under uncertainty
- average cost
- factored mdps
- action space
- partially observable markov decision processes
- reward function
- action sets
- reachability analysis
- semi markov decision processes
- model based reinforcement learning
- average reward
- state abstraction
- state and action spaces
- search space
- multi agent