4G電信產業已經在國外行之有年,但台灣卻是屬於一個新興的產業,加上國人對於行動網路和資料隨時隨地傳輸的需求量也跟著成長。本研究最主要的目的是希望能夠預測4G電信產業的在台灣未來的發展趨勢,另外台灣4G產業在2014年5月正式提供服務,過去並沒有相關的歷史統計資料。 在經過一年的資料蒐集後,在資料量有限的強況下,我決定用灰色理論當作我的預測工作,GM(1,1)預測模型在灰色理論系統中相當知名,其最重要的特就是在資料量不足時能以不錯的預測準確度和預測能力。 在研究期間,發現到GM(1,1) 在某些特定的資料成長情況下,會產生資料發散的情況,因此在不斷的嘗試和研究相關的玩現後,我決定採用指數平滑法來嘗試解決此問題,最後成功的將指數平滑法和GM(1,1)模型做結合,發展出一套新的預測模型,也成功的解決GM(1,1)在某些特定資料所產生的資料發散問題。
4G telecommunication industry has been operating for years in foreign countries, but in Taiwan it is an emerging industry in the beginning. And the people who demand for mobile network for data transmission are also growing. The main purpose of this study is to forecast the trend of 4G telecommunication industry in Taiwan. The 4G telecommunications of Taiwan began providing services in May 2014, and there are no related historical statistics. After collecting data for a year, in the limited amount of data, I decided to use the Grey Theory as the forecasting tool for this research. The GM (1,1) model is very famous in grey system theory, and it has successfully been widely applied in various research fields. The most important characteristic of GM (1,1) is that it features good forecasting results while the current data is insufficient in the related studies. During the research, I found that GM (1,1) model was not fully applicable to the data. In addition, we find the problem about data divergence of the forecast results. In order to solve this problem, I successfully combined the Holt’s two parameter exponential model and GM (1,1) model. Finally, I get the better accurate and significant achievements.