Evolutionary reinforcement learning of dynamical large deviations.
Stephen WhitelamDaniel JacobsonIsaac TamblynPublished in: CoRR (2019)
Keyphrases
- large deviations
- reinforcement learning
- state dependent
- optimal policy
- importance sampling
- queueing systems
- heavy tailed
- genetic algorithm
- state space
- control problems
- markov processes
- queue length
- optimal control
- dynamic programming
- learning algorithm
- mathematical programming
- asymptotically optimal
- generalization bounds
- machine learning