淡江大學機構典藏:Item 987654321/105701
English  |  正體中文  |  简体中文  |  Items with full text/Total items : 62796/95837 (66%)
Visitors : 3641975      Online Users : 560
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/105701


    Title: 以決策樹結合領域導向方法挖掘不可預期模式
    Other Titles: Mining unexpected patterns using decision tree with domain driven approach
    Authors: 詹千慧;Chan, Chien-Hui
    Contributors: 淡江大學資訊工程學系博士班
    蔣璿東
    Keywords: 不可預期模式;領域導向資料探勘;決策樹;治療比較;Unexpected pattern;domain-driven data mining;Decision tree;Treatment comparison
    Date: 2015
    Issue Date: 2016-01-22 15:03:03 (UTC+8)
    Abstract: 不可預期模式的有趣之處在於,它們與人的既有知識相悖或是出乎意料,所以可能可以提供一些不同的觀點給研究人員參考,並可用以對未來研究的內容與方向提出建議;因此,本研究將提出一個不可預期模式探勘模型,以找出與領域專家前導知識相違背之不可預期模式。傳統資料探勘的過程是強調以資料為中心的模式探勘,環境、人類經驗等等的因素經常是被過濾或是大量的簡化,較不考慮個別使用者的需求或是領域相關的知識。在本研究中是使用醫學上的經陰道超音波引導抽取術之追蹤資料進行分析,由於臨床研究的環境因素較為複雜,所以發展一個可以與使用者互動並將領域前導知識、領域限制及專家知識介入資料探勘過程的模型是很重要的。同時,因醫學資料中經常包含大量數值變數,而決策樹可以同時處理數值及類別型資料,本研究所提出之模型使用決策樹結合領域導向資料探勘中封閉迴圈、深入探勘的概念,比較不同治療方式的治癒率並挖掘不可預期模式。
    Unexpected patterns are interesting because they are contrast with the prior knowledge or unexpected. Therefore, unexpected patterns may provide researchers with different vision for future research. In this study, we propose an unexpected pattern mining model to find patterns that contrast with the prior knowledge of domain users. Traditional data mining emphasizes data-centered mining for interesting patterns. During the data mining process, environmental factors are usually filtered or simplified. Individual user requirements and domain-related knowledge are less considered. In this study, we use retrospective data from transvaginal ultrasound-guided aspirations to conduct our analysis. Since clinical studies are conducted in complex environments, we believe that it is important to develop an interactive mining model that involves prior domain knowledge, constraints, and expert knowledge. Meanwhile, medical data usually contain plenty of continuous variables. Decision tree algorithms can deal with both continuous and categorical variables at same time. Therefore, the proposed model uses decision trees to compare the recovery rates of two different treatments. By applying the concept of domain-driven data mining, we repeatedly utilize decision trees in a closed-loop, in-depth mining process to find unexpected and interesting patterns.
    Appears in Collections:[Graduate Institute & Department of Computer Science and Information Engineering] Thesis

    Files in This Item:

    File Description SizeFormat
    index.html0KbHTML196View/Open

    All items in 機構典藏 are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback