Approximate Linear Programming and Decentralized Policy Improvement in Cooperative Multi-agent Markov Decision Processes.
Lakshmi MandalChandrashekar LakshminarayananShalabh BhatnagarPublished in: CoRR (2023)
Keyphrases
- markov decision processes
- linear programming
- optimal policy
- policy evaluation
- policy iteration
- dynamic programming
- cooperative multi agent
- average cost
- reinforcement learning
- factored mdps
- markov decision process
- infinite horizon
- decentralized control
- state space
- finite horizon
- average reward
- partially observable
- markov decision problems
- state and action spaces
- finite state
- action space
- reward function
- linear program
- decision processes
- planning under uncertainty
- dec pomdps
- least squares
- decision problems
- multi agent
- transition matrices
- expected reward
- decision theoretic planning
- total reward
- logic programming
- discounted reward
- multistage
- reachability analysis
- continuous state spaces
- stationary policies
- control policies
- reinforcement learning algorithms
- markov games
- long run
- model based reinforcement learning
- partially observable markov decision processes
- search algorithm
- objective function
- temporal difference
- action sets
- optimal solution
- inventory level
- np hard