Analyses about efficiency of reinforcement learning to supply chain ordering management

The Reinforcement Learning (RL) is an efficient machine learning method for solving problems that an agent has no knowledge about the environment a priori. Improving efficiency of decision-making practices in a supply chain is a major competitive domain in today´s uncertain business environments. The bullwhip effect is an important phenomenon in the supply chain, in which the order variability increases as moving up along the supply chain. This paper proposes a multiagent coordination mechanism utilizing RL method to the supply chain ordering management. Further, the analyses about the efficiency of the method are discussed in detail based on some representative test data. Results show that the RL agent reduces the bullwhip effect efficiently in the stochastic supply chain.