Sufficiency of Deterministic Policies for Atomless Discounted and Uniformly Absorbing MDPs with Multiple Criteria.
Eugene A. FeinbergAlexey B. PiunovskiyPublished in: SIAM J. Control. Optim. (2019)
Keyphrases
- multiple criteria
- optimal policy
- decision problems
- markov decision processes
- stationary policies
- finite horizon
- markov decision process
- infinite horizon
- state space
- average reward
- average cost
- dynamic programming
- influence diagrams
- reinforcement learning
- discount factor
- markov chain
- markov decision problems
- finite state
- reward function
- policy iteration
- discounted reward
- multi criteria
- decision aid
- sufficient conditions
- decision makers
- total reward
- partially observable markov decision processes
- long run
- initial state
- group decision making
- multi attribute
- partially observable
- multi objective
- mathematical programming
- preference models
- artificial intelligence
- model free
- neural network
- policy search
- message passing
- evolutionary algorithm