資料探勘應用於呆滯料預警機制之建立; A Establish on the Application of Data Mining on Surplus Warning Mechanism

NCUIR > College of Electrical Engineering & Computer Science > Executive Master of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/61024

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/61024

Title:	資料探勘應用於呆滯料預警機制之建立;A Establish on the Application of Data Mining on Surplus Warning Mechanism
Authors:	潘陳世杰;Pan-Chen,Shih-Jie
Contributors:	資訊工程學系在職專班
Keywords:	資料探勘;類神經網路;庫存;羅吉斯迴歸;呆滯料;Data Mining;Artificial Neural Network;Stock;Logistic Regression;Surplus
Date:	2013-07-03
Issue Date:	2013-08-22 12:10:01 (UTC+8)
Publisher:	國立中央大學
Abstract:	由於偏光板銷售市場逐年衰退，在製造端獲利減少的狀況下，將目標轉向其他地區有需求的銷售市場，並努力去化庫存的呆滯料，以減少企業在庫存資金累積的壓力，進而提升企業在同業間的競爭力。庫存成品變成呆滯料之原因，會隨著使用年限的增加而變高。但未超過用期限之內的成品，通常能藉著促銷或是折扣之方式進行銷售，避免為呆滯料。所以如何能有效的預測成為呆滯料之成品，使行銷 / 生管人員能提前進行促銷或是進行備料之調整，便成了此次研究想探討的問題。本研究是採用X公司DW系統作為資料的來源，利用資料探勘技術，以SQL Server 2008 R2為建立預測模型工具，並使用類神經網路與羅吉斯迴歸探勘技術建立呆滯料預測模型，管理者可以根據預測模型的結果，及早採取對應措施，進而降低呆滯料發生機率，可作為管理者後續降低庫存資金之參考。根據研究結果得知，本研究所建立的預警機制，可以提供決策者後續降低庫存參考依據與減少公司報廢金額的損失。 Due to the decline of polarizer sales market year by year and decreasing manufacturing profits, enterprises are turning to other sales markets with needs. This is an effort to digest stock materials to reduce the pressure of surplus fund accumulation. Thus, this can enhance the competitiveness of enterprises in the same industry. The reason stock products are turning to surplus materials and the usage period is proportional to this increase. Unexpired products are sold at discounted prices to avoid turning to surplus materials. Hence, the purpose of this study is focus on how to efficiently predict the products turning to surplus materials to enable marketing and production managers to prepare to promote or to adjust the materials. In this study,”X” company’s DW system is used as a source. We use SQL Server 2008 R2, a kind of technology of data mining, to establish the prediction model, aided by of artificial neural network and logistic regression technology to establish the prediction model of surplus materials. According to the results of predictive models, managers can take contingency measures to lower the probability of surplus materials. In the future, managers can use this study to reduce the expense of the stock fund. Early warning mechanisms established in this study can provide the reference for decision-makers to reduce stock and help companies reduce the loss of scrapped funds.
Appears in Collections:	[Executive Master of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	1116	View/Open

社群 sharing

Loading...