Capacity planning and interactive dynamics for motherboard

• Main Authors: Yun, Ying, 雲穎
• Other Authors: Wang, Chih-Hsuan
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/359tut

碩士 === 國立交通大學 === 工業工程與管理系所 === 107 === Motherboard is known as the core of computer systems. Under the fierce competition in the industry, more and more features are replaceable, which causes crowding and competition effect. Along with the shrink of desktop and laptop market in the world, motherboard industry will certainly be affected to impact on Taiwan’s overall manufacturing industries. This research will discuss three major parts which matters: capacity planning, revenue forecasting, and dynamic interaction. Three metrics, such as RMSE, MAD, and MAPE, are used in this research to measure the quality of the model. Time Series and machine learning are adapted to forecast the demand of motherboard (global shipments) as well as perform capacity planning for motherboard. Taiwanese computer companies used in this research include ASUS, MSI, ASROCK, and GIGABYTE. In the past, the variables which affect the sales revenues were more complicated than those affect the global shipments, which can be trade variables, consumer products, economic factors, etc. This research applies the concept of product mix to forecast the sales revenue of computer companies. As we know, the motherboard requires a large number of chips, which enables domestic manufacturers to actively develop related components and then indirectly drives the growth of Taiwan semiconductor industry, especially driven by the motherboard industry and related computer products including desktops, laptops, and servers. According to this fact, we try to explore the dynamics between the motherboard market and the market of the consumer products. In particular, Lotka- Volterra competition model has been developed to address the relationship between them and reveal managerial insights.