抄録
In recent years, diversified market needs have led to the constant use of high-mix, low-volume production in the manufacturing industry. However, there has been little research on the development of Demand Forecasting models that appropriately consider everyday market needs or of product-mix decision making methodologies that strategically use such models.
In this study, a new Demand Forecasting model was developed based on genetic programming (GP), using experimental forecasting factors derived from sales activities with customers in metropolitan areas as the decision variable. The authors identify the process used to achieve improved Demand Forecasting accuracy by solving product-mix decision-making and how it impacts the manufacturing industry using Profit Contribution Analysis. Furthermore, a method to quantitatively identify the need for introducing new technologies to realize product-mixes using the aforementioned method is examined. Finally, the effectiveness of this method is validated using actual data from small and medium-sized manufacturers located in a regional area.
