Budget Allocation using Weakly Coupled, Constrained Markov Decision Processes.
Craig BoutilierTyler LuPublished in: UAI (2016)
Keyphrases
- markov decision processes
- state space
- finite state
- reinforcement learning
- policy iteration
- optimal policy
- decision theoretic planning
- risk sensitive
- finite horizon
- transition matrices
- planning under uncertainty
- partially observable
- dynamic programming
- factored mdps
- markov decision process
- reachability analysis
- model based reinforcement learning
- average cost
- decision processes
- average reward
- reinforcement learning algorithms
- action space
- reward function
- learning algorithm
- state and action spaces
- action sets
- heuristic search