Learning in infinite-horizon inventory competition with total demand observations.

Ashkan Zeinalzadeh Aydin Alptekinoglu Gurdal Arslan

Published in: ACC (2012)

Keyphrases

infinite horizon
stochastic demand
lead time
long run
single item
optimal control

holding cost
learning algorithm
optimal policy
dynamic programming
finite horizon
inventory level

inventory costs
lost sales
production capacity
reinforcement learning
inventory policy
periodic review

partially observable
markov decision processes
optimal production
search algorithm
demand distributions