A Deep Reinforcement Learning Architecture for Multi-stage Optimal Control.

Published in: CoRR (2019)

Keyphrases

optimal control
multistage
dynamic programming
reinforcement learning
control problems
assembly systems
optimal policy
feedback control
single stage

stochastic programming
optimal control problems
production system
state space
infinite horizon
class of nonlinear systems
stochastic optimization
control strategy

markov decision processes
lot sizing
real time
control law
function approximation
average cost
actor critic
rl algorithms

policy gradient
linear program
lot streaming
inventory systems
stochastic control
linear quadratic
knapsack problem
learning algorithm