The Analysis of Grey GM(1,1) Model for Discrete Data

碩士 === 清雲科技大學 === 機械工程研究所 === 93 === Calculation by using similar random distribution series and functional values, it is found that the prediction accuracy of three alpha GM(1,1) in four points rolling models, is better than traditional grey system GM(1,1) models. However, it is worse in functional...

Full description

Bibliographic Details
Main Authors: Han-Wei Kuo, 郭漢偉
Other Authors: Ching-China Ji
Format: Others
Language:zh-TW
Published: 2005
Online Access:http://ndltd.ncl.edu.tw/handle/61409882204374352773
Description
Summary:碩士 === 清雲科技大學 === 機械工程研究所 === 93 === Calculation by using similar random distribution series and functional values, it is found that the prediction accuracy of three alpha GM(1,1) in four points rolling models, is better than traditional grey system GM(1,1) models. However, it is worse in functional value prediction. Due to over accumulating generation, it will make the rule of exponential getting grey, such that those two models are still not good enough to predict the cross point of functional values. Only traditional GM(1,1) model is more effective to foresee the cross point of 20 weeks moving average line and 120 weeks moving average line. The trend of moving average line can be judged by Granvile eight rules; it helps to forecast the trend when we can understand the tendency of moving average line in advance.