State-Continuity Approximation of Markov Decision Processes via Finite Element Analysis for Autonomous System Planning.
Junhong XuKai YinLantao LiuPublished in: CoRR (2019)
Keyphrases
- markov decision processes
- finite element analysis
- state space
- partially observable
- planning under uncertainty
- decision theoretic planning
- action space
- optimal policy
- finite state
- heuristic search
- finite element
- macro actions
- dynamic programming
- markov decision process
- planning problems
- reinforcement learning
- state abstraction
- transition matrices
- reinforcement learning algorithms
- reachability analysis
- real time dynamic programming
- policy iteration
- average cost
- discounted reward
- infinite horizon
- decentralized control
- ai planning
- model based reinforcement learning
- computer aided design
- probabilistic planning
- state variables
- partially observable markov decision processes
- initial state
- markov decision problems
- decision theoretic
- average reward
- data model
- machine learning