Asymptotic Security by Model-based Incident Handlers for Markov Decision Processes.
Hampei SasaharaHenrik SandbergPublished in: CoRR (2021)
Keyphrases
- markov decision processes
- optimal policy
- interval estimation
- finite state
- state space
- transition matrices
- dynamic programming
- decision theoretic planning
- reinforcement learning
- decision processes
- model based reinforcement learning
- security issues
- reachability analysis
- average cost
- markov decision process
- policy iteration
- risk sensitive
- average reward
- infinite horizon
- semi markov decision processes
- planning under uncertainty
- finite horizon
- partially observable
- factored mdps
- information security
- model free
- action space
- state abstraction
- reinforcement learning algorithms
- reward function
- np hard
- machine learning