透過您的圖書館登入
IP:18.191.195.110
  • 期刊

應用關聯規則探勘台灣地區果品批發市場行情資料庫之研究

A Study of Applying Association Rules in the Trading Report Database of Fruit Wholesale Markets in Taiwan

摘要


果品市場行情資料庫中,包含大量的果品種類、平均價格及交易量等批發交易資料,若能深入處理與分析,可找出隱藏在資料庫中有用的資訊和知識,作為作業控制或經營決策參考。本文在不限定果品品項並保留品種資料之前提下,針對台灣地區消費地型態之果品批發市場交易行情資料庫進行資料探勘,利用關聯規則尋找果品之間的相關性,探勘出大量的關聯規則後,再運用設立的判別標準進行刪除,以尋找出有趣的關聯規則。 研究結果發現,並非所有同名不同種果品,會產生相同事件。部分非直接食用之果品關係被忽略;而部分果品之間呈現高信賴度,代表果品之間其有關聯,建議農政單位可以針對具有關聯性的果品配合其他研究方法深入研究,瞭解果品之間的確切關聯,以建立果品預警系統。

並列摘要


B sod on the condition I unlimited fruit items and preserved species of fruit, this research uses association rules to find possible associations among fruits in the Trading Reports Database of Fruit Wholesale Markets in Taiwan. First, we transfer the fruits' daily price, trading volume, the price change and the trading volume change to user-defined intervals and then use association rules to find the correlation of fruits. After mining the various association rules, we establish differentiated criteria to do deletion o as to detect any association rule. In the results, we find that some fruit of the same genus but different species have different events and some indirect edible fruits correlation can be ignored. The high confidence means that a correlation exists among these fruits. Thus, we suggest the agricultural institution uses other research method to confirm this correlation to establish the fruit early warning system.

並列關鍵字

Fruit Wholesale Market Data Mining Association Rule

被引用紀錄


王國光(2006)。西螺農產品批發市場之吸引力-動態系統研究〔碩士論文,國立虎尾科技大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0028-1501201314421093

延伸閱讀