Exploiting Anonymity in Approximate Linear Programming: Scaling to Large Multiagent MDPs.
Philipp RobbelFrans A. OliehoekMykel J. KochenderferPublished in: AAAI Fall Symposia (2015)
Keyphrases
- linear programming
- multi agent
- markov decision problems
- factored mdps
- dec pomdps
- planning under uncertainty
- markov decision processes
- dynamic programming
- linear program
- reinforcement learning
- factored markov decision processes
- cooperative
- average cost
- policy iteration
- multiagent systems
- state space
- integer programming
- policy evaluation
- optimal solution
- decision theoretic planning
- objective function
- np hard
- single agent
- column generation
- quality guarantees
- reward function
- privacy protection
- constraint propagation
- policy search
- intelligent agents
- multiple agents
- decision theoretic
- privacy preserving
- optimal policy