The knowledge sharing model on supply chain simulation using recurrent neural network

The bullwhip effect is one of the most important problems on the supply chain management. It is a phenomenon that fluctuation in demands increases on upstream of supply chain. Inaccurate demand forecasting accompanying lack of communication is well known as a cause of bullwhip effect. Nevertheless, there is almost no sufficient argument on the influence of the bullwhip effect on demand forecasting through communication with business contacts. In this paper, we propose the framework and simulation model of supply chain system based on demand forecasting through communication with business contact. Here, the simulation model was constructed by modeling about the past transaction data as knowledge sharing with business contacts´ company. Recurrent neural network that is excellent in time-series prediction was used for demand forecasting in this simulation. We confirm the effectiveness of our proposal through comparative experiments using inventory management simulation on conventional models and on proposed model.