1 paper
Lucain Pouget, Timo Hasenbichler, Jakob Auer +2
We investigate the feasibility of deploying Deep-Q based deep reinforcement learning agents to job-shop scheduling problems in the context of modular production facilities, using d…