Abstract
We develop a modified version of deep deterministic policy gradient(DDPG) algorithm, which is one of the most popular deep reinforcement learning algorithms dealing with continuous action spaces, and apply it to the infinite-horizon newsvendor problems with constant lead time. Reflecting the key features of the inventory management problems, our algorithm, named as Inventory based DDPG (IDDPG), is differentiated in the state variables, action spaces, and cost functions compared to the DDPG. By conducting numerical experiments and comparing the performances, we found that when applied to the inventory management problems, IDDPG is able to find the near-optimal solutions and outperforms the DDPG in most cases. We also found that our IDDPG can be applied to a inventory management problem whose optimal solution is not known.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Journal of the Korean Society of Supply Chain Management
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
