Efficient algorithms for Risk-Sensitive Markov Decision Processes with limited budget.
Daniel Augusto de Melo MoreiraKarina Valdivia DelgadoLeliane Nunes de BarrosDenis Deratani MauáPublished in: Int. J. Approx. Reason. (2021)
Keyphrases
- multi agent
- risk sensitive
- markov decision processes
- limited budget
- reinforcement learning
- optimal policy
- state space
- finite state
- average cost
- reinforcement learning algorithms
- infinite horizon
- policy iteration
- dynamic programming
- partially observable
- planning under uncertainty
- reward function
- finite horizon
- action space
- decision processes
- average reward
- partially observable markov decision processes