查詢結果分析
來源資料
頁籤選單縮合
題名 | 使用灰色理論於改善資料品質之研究=A Study of Using Grey Theory in Improving Data Quality |
---|---|
作 者 | 夏自立; 龔榮源; 林佳姿; | 書刊名 | 遠東學報 |
卷期 | 23:1 民95.03 |
頁次 | 頁211-220 |
分類號 | 312.13 |
關鍵詞 | 灰色理論; 資料探勘; 缺漏資料; 多維度資料; Grey theory; Data mining; Missing data; Multi-dimensionality data; |
語文 | 中文(Chinese) |
中文摘要 | 資料探勘是一種能從大量資料中挖掘出有用資訊的技術,原始資料品質的良 窳攸關資料探勘所發掘「知識」的正確性與代表性。資料品質不佳是資料庫 中普遍存在的問題,缺漏資料過多與資料過於複雜則是影響資料品質的重要 因素之一。針對如何改善資料品質以提昇資料探勘成效此項議題,本研究以 灰色系統理論為基礎,利用灰生成技術來探討其在缺漏資料填補上的應用; 另運用灰關聯分析技術來萃取多維度資料間的潛在關聯性特徵,以解決資料 過於複雜的問題。本研究經由模式推導與範例資料的初步驗證,發現灰生成 能有效的處理時間數列資料缺漏的問題,而灰關聯分析在特定問題下具有簡 化資料複雜性與特徵萃取的能力。 |
英文摘要 | Data mining is one technique that can discover useful information from large amount of data. The quality of raw data is essential to the accuracy and representation of knowledge acquired through data mining. However, the poverty of data quality often exists in a variety of databases, and plenty of missing data as well as data complexity has much influence upon data quality. Accordingly, in this paper we propose new methods, which are based on the grey system theory, to improve performances of missing data preprocessing and multi-dimensionality data transformation. To complete missing data, we apply the grey generating technique to filling the missing values. As to treatment of multi-dimensionality data, the grey relational analysis is utilized to extract potential relevant terms from raw data. Some illustrative examples are also given to demonstrate the proposed methods. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。