| Summary: | 碩士 === 僑光技術學院 === 管理研究所 === 94 === Abstract
According to the year-end statistics published by The Fair Trade Commission in 2005, the accumulative numbers of franchised convenient stores in Taiwan have reached a figure of 8,892 stores; Making Taiwan the nation with the highest convenient store density per person worldwide. Currently, on average, for every 2556 habitants in Taiwan, there is one corresponding convenient store.
Stops at convenient stores have become a daily routine for most Taiwan residents. Hence getting a grip of this huge market has become a major focal point for all cold beverages manufacturers. And this is where Neutral Network comes into play.
Neutral Network is one of the most precise predicting system that has been gaining popularity amongst various forecasting systems in the recent years. In the past it was mainly utilized for the prediction of electricity and financial trends, but not commonly used for the forecasting of short term product turnover cycles.
The Aim of this study is to design a forecasting system based on Matlab software that utilizes the excellent forecasting characteristics of BPN. This combination of the powerful Matlab software along with the outstanding forecasting attributes of Back-Propagation Network shall constitute a reliable forecasting system that will help cold beverages makers in anticipating the short term product turnover cycles of its products, elevating the corporate competitiveness advantage in the market.
|
|---|
