Digital Twins Composition via Markov Decision Processes.
Giuseppe De GiacomoMarco FavoritoFrancesco LeottaMassimo MecellaLuciana SiloPublished in: ITBPM@BPM (2021)
Keyphrases
- markov decision processes
- finite state
- optimal policy
- reinforcement learning
- policy iteration
- state space
- transition matrices
- decision theoretic planning
- factored mdps
- reinforcement learning algorithms
- decision processes
- finite horizon
- dynamic programming
- model based reinforcement learning
- reachability analysis
- risk sensitive
- infinite horizon
- planning under uncertainty
- decision diagrams
- action space
- markov decision process
- average cost
- real time dynamic programming
- state and action spaces
- multistage