透過您的圖書館登入
IP:18.218.127.141
  • 學位論文

羽絨產業原物料價格預測模式之研究

Modeling and Forecasting the Raw Material Price in Feather Industry

指導教授 : 吳忠敏
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


近年來由於氣候異常的問題越來越嚴重,導致農業生產量重挫,國際市場原物料發生供需異常、價格波動頻繁的現象,企業在規劃生產成本上的困難度大增,預測原物料的價格成為所有企業不容克緩的議題。而在近年「千年寒冬」的議題下,羽絨產品藉由其輕量與保暖等等特點成為近年來最主要的禦寒商品,但台灣由於地狹人稠,羽絨原物料仰賴進口,因此若能提早預測國際市場之羽絨價格,則可協助台灣羽絨企業在生產規劃階段找到物廉價美之原物料。 本研究以近年來因寒冬備受矚目的羽絨製品為本研究的研究標的,利用歷年來文獻找出影響羽絨產業原物料價格之因素,並且利用相關分析找出影響羽絨產業原物料相關之因素,再利用複迴歸分析、時間序列分析、個案推理各別建立台灣主要進口國之羽絨產業原物料價格的預測模型,以期找出最佳之預測模式,並以向量自我迴歸模型觀察變數間之影響關係。 研究結果顯示在利用複迴歸得到的所有預測模型中,最高的解釋度達72.3%;利用;時間序列模型之最高解釋度達75.7%;並提供兩種個案推理之結果,提供相關產業做為參考。

關鍵字

預測 原物料價格 羽絨

並列摘要


Recently, the problem of climate has been more and more serious. So the output of crop has been decreased. The supply and demand of raw materials can’t balance, so the fluctuation in materials price also been more and more violent. Companies can’t estimate their cost simply, forecasting the price of raw materials goes more and more important. With the colder winter, the products of feather which light and keep warm has been the most important product of keep warm. But the feather businesses in Taiwan have to get the raw material from oversea. If we can forecast the feather’s price of importing country, that will be a good help in finding material which cheap and fine. This paper uses the feather and down which is popular item at cold winter as the object. Considering past studies viewpoint and using the correlation analysis to find the relative factor. Using multiple regression analysis, time-series analysis, and case-based reasoning construct the models of forecasting price of feather and down. Aiming to find the best forecasting model of feather’s price. Alos use vector autoregression to observe the relationship of variables. First, the results indicate that the highest at all proposed models in multiple regression analysis explained up to 72.3% of the variance from the regression as the response rate. Second, the highest at all proposed models in time-series analysis explained up to 75.7% of the variance from the regression as the response rate. Last, provide two results of case-based reasoning. This analysis can provide enterprise as reference.

參考文獻


[34] 楊金聲,利用類神經網路與線性迴歸進行成本預測之研究-以印刷電路板產業為例,碩士論文,中原大學資訊管理系,桃園,2005。
[29] 程惟國,類神經網路應用於颱風期間全台區域農業損失之研究,碩士論文,國立臺灣大學生物資源暨農學院生物環境系統工程學研究所,台北,2010。
[20] 唐琦、徐森雄,「臺灣南部地區農業氣象環境與災害發生潛勢」,作物、環境與生物資訊,第4期,2007,第11-22頁。
[18] 林師模、張彩姿、鄰晉勗、翁永和,「能源及原物料價格上漲之跨國傳遞效果」,臺灣經濟預測與政,第40期,2010,第1-41頁。
[17] 林茂文,時間序列分析與預測:管理與財經之應用,第三版,台北:華泰文化事業股份有限公司,2006。

延伸閱讀