Application of Grey Theory to Forecast Short-Term Sales of Taiwan Fastener Industry

Abstract

Recent developments in the fastener industry has heightened interest in the relationship between data mining and short-term sales forecasting Previous studies have reported inventory management problems in Taiwan’s fastener industry However important questions remain to be resolved including how to improve the accuracy and reliability of sales forecasting when dealing with short-term limited data The grey model is proposed to overcome this problem through accumulated generating operations (AGO) The background values represented as coefficient ? play a critical role that determines and affects the precision of the GM model The purpose of this research is to improve the accuracy and reliability of the grey prediction model GM (1 1) with limited data in the earlier part of the prediction process Therefore this study proposes a modified GM (1 1) based ant colony system (ACS) to obtain the optimal ? values Meanwhile two kinds of data are employed: the Synthetic Control Chart Time Series dataset (SCCTS) from the Knowledge Discovery Database (KDD) and the fastener industry statistics from Taiwan Customs Import and Export data in order to verify the effectiveness of the ACSGM (1 1) model It is compared with two other models GM (1 1) and AGM (1 1) The results reveal that ACSGM (1 1) significantly improves accuracy The findings may serve as a guide in strategic planning for support management decisions related to forecasting demand for case companies and for the fastener industry

Date of Award	2018 Feb 22
Original language	English
Supervisor	Chinho Lin (Supervisor)