科技擴散與替代模型之研究-以晶圓製程技術為例

Abstract

雖然科技暨產品擴散模型已廣泛地被研究並應用在市場行銷與預測上，有關多世代科技暨產品擴散替代模型的研究仍相當有限，而傳統多世代科技暨產品擴散替代模型的基本假設–新科技或產品會完全取代舊科技或產品，也與實證資料不符。本研究的目的為結合基因演算法與數值方法–已被驗證比傳統統計方法有更佳的參數擬合效果，來對新提出的多世代科技暨產品擴散替代模型作參數擬合。在晶圓代工的實證資料上，本研究提出的多世代科技暨產品擴散替代模型比傳統多世代科技暨產品擴散替代模型有更好的擬合效果，進而由參數擬合結果可更了解不同世代科技暨產品與擴散替代模型間的管理意含。本研究結果亦可應用在策略規劃與市場預測上。

Researches of technology or product diffusion models have been commonly studies and applied in marketing research and forecasting. However, researches on multiple generation diffusion models are critically limited, and empirical data can challenge the basic assumption of traditional multiple generation diffusion models: new-generation technologies will completely replace old technologies. The aim of this study is to present execution of Genetic Algorithm plus numerical method, which is verified better than traditional ordinary least square methods in fitting accuracy, in the proposed diffusion and substitution model of multiple generation technologies. Compared with traditional multiple generation model in empirical data of Semiconductor Foundry Industry, the proposed model reveals better fitting results. To assess the consequences of parameters fitting results of the proposed model, this study may lead to a better understanding of relationships between diffusion models and management interpretations. The result of this study could also be useful to researchers responsible for strategy planning and market forecasting.