Adversarial Reinforcement Learning under Partial Observability in Software-Defined Networking.
Yi HanDavid HubczenkoPaul MontagueOlivier Y. de VelTamas AbrahamBenjamin I. P. RubinsteinChristopher LeckieTansu AlpcanSarah M. ErfaniPublished in: CoRR (2019)
Keyphrases
- partial observability
- reinforcement learning
- partially observable
- symbolic model checking
- belief state
- planning problems
- markov decision process
- belief space
- fully observable
- state space
- markov decision processes
- multi agent
- learning agent
- machine learning
- reinforcement learning algorithms
- partially observable markov decision processes
- decision problems
- knowledge acquisition
- function approximation
- model checking
- orders of magnitude
- optimal policy
- search space
- computational complexity
- hidden state
- bayesian networks
- learning algorithm