SPSC: a new execution policy for exploring discrete-time stochastic simulations.
Yu-Lin HuangGildas MorvanFrédéric PichonDavid MercierPublished in: CoRR (2019)
Keyphrases
- markov processes
- optimal control problems
- control policies
- markov chain
- state dependent
- optimal control
- optimal policy
- numerical simulations
- simulation environment
- model free reinforcement learning
- steady state
- monte carlo
- simulation models
- finite state
- production planning
- expected cost
- average cost
- stochastic processes
- stochastic optimization
- markov decision processes
- stochastic control
- learning algorithm