Application of a Grey Forecasting Model to Predict the Number of Subscribers of Multimedia Service Platform

• Main Authors: Ho, Ming-Hsun, 何明訓
• Other Authors: Chu, Ching-Wu
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/84644134613891099528

碩士 === 國立臺灣海洋大學 === 航運管理學系 === 102 === Forecast can often provide effective forward-looking information in the decision-making process. So accurate forecasting is an important issue for business sustainability. In this study, a GM(1,1) model is constructed to predict the number of subscribers of a multimedia service platform based on the information on the National Communications Commission (NCC) website. The monthly data of the number of subscribers were collected from year 2009 to 2013. Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) are used to compare the forecast accuracy of GM(1,1) models with different data. After fitting the original data, we found that the GM (1,1) model is suitable for forecasting the number of subscribers of a multimedia service platform. Based on Mean Absolute Percentage Error, the GM (1,1) model with four data provides the lowest forecast error of 2.7%, and GM (1,1) models with five and with six data also provide a high-precision forecast that the MAPE is under 10%. Our proposed forecast model can provide references to Communications authorities in planning digital TV content, forecasting television digitization popularity and to enterprises in planning multimedia products, setting the strategic direction and deciding the order quantity and inventory level of digital Set-top boxes.