請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/41440
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 陳正剛 | |
dc.contributor.author | Chien-Ming Chen | en |
dc.contributor.author | 陳建名 | zh_TW |
dc.date.accessioned | 2021-06-15T00:19:15Z | - |
dc.date.available | 2019-09-06 | |
dc.date.copyright | 2009-02-18 | |
dc.date.issued | 2009 | |
dc.date.submitted | 2009-02-13 | |
dc.identifier.citation | [1] Breiman L., Friedman J. H., Olshen R. A. and Stone C. J., “Classification and Regression Trees”, Monterey, California: Wadsworth and Brooks/Cole, Belmont, 1984.
[2] T. R. Ho, “Sample-Efficient Regression Tree for Binary and Ordinal Attributes and Continuous Target”, M.S. Thesis, Graduate Institute of Industrial Engineering, National Taiwan University, 2003. [3] C.W. Liu, “Enhanced Sample-Efficient Regression Trees with MaxF Selection Criterion and Attribute Combination Selection”, M.S. Thesis, Graduate Institute of Industrial Engineering, National Taiwan University, 2004. [4] Y.X. Huang, “Sample-Efficient Regression Trees for Attributes with Maxed Continuous and Discrete Effects – A Piecewise-Linear Regression Tree”, M.S. Thesis, Graduate Institute of Industrial Engineering, National Taiwan University, 2005. [5] Gilbert Strang, “Linear Algebra and its application”, Belmont, CA: THOMSON and Brooks/Cole, pp174-185, 2006 [6] Yu-Wei Lin, “Gram-Schmidt Transformation Minimization Algorithm and Its Applications to Regression Analysis with Multicollinearity” , M.S. Thesis, Graduate Institute of Industrial Engineering, National Taiwan University, 2007. [7] G. W. Stewart “Collinearity and least squares regression”, Statistic Science, 1987 | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/41440 | - |
dc.description.abstract | 分類迴歸樹 (classification and regression tree, CART) 在資料採擷(data mining)裡是很常被使用的方法,透過對樣本資料的切割,將觀察值以二元的方式進行分類。但CART會隨著資料的分割,使樣本數快速的減少,減低預測的可靠性,因此提高樣本使用效率的迴歸樹 (Sample Efficient Regression Tree, SERT) 便提出來,利用連動效力檢定 (Interaction Effect Test) 來避免不必要的分割,但是,對於樣本裡存在著二次效力 (Quadratic Effect) 時,CART和SERT都無法檢測出來,因此我們提出了提高樣本使用效力的逐段二次迴歸樹(Sample-Efficient Piecewise Quadratic Regression Tree)來解決這個問題。
首先,我們發展了一種逐段二次迴歸模型(Piecewise Quadratic Regression Model)來對樣本中存在的二次效力作檢測,接著還利用Gram-Schmidt Process的手法發展出選擇變數的一次項和二次項的方法,以避免因為兩者間的高度共線性而造成變數選取的錯誤。 最後,我們列舉了十七個不同類型的模擬方案以及實際的例子來驗證我們提出的方法。 | zh_TW |
dc.description.abstract | The classification and regression tree (CART) is a popular method in data mining. It classifies the responses by sequentially splitting the sample into two branches. In CART, the sample size will deplete quickly and the reliability of prediction will diminish with splitting sample. Therefore, the sample efficient regression tree (SERT) is proposed. It uses interaction effect test to avoid unnecessary splits. However, CART and SERT can not detect the quadratic effect. For this reason, we propose the sample-efficient piecewise quadratic regression tree to solve the problem.
First we develop the piecewise quadratic regression model to detect the quadratic effect. Then, we use the Gram-Schmidt to resolve the possible multicollinearity issue between the linear effect and the quadratic effect. With this process, we can avoid wrong attribute selection resulted from statistical insignificance due to collinearity between the linear effect and the quadratic effect. Finally, we use seventeen simulated cases and a real case to verify our proposed method. | en |
dc.description.provenance | Made available in DSpace on 2021-06-15T00:19:15Z (GMT). No. of bitstreams: 1 ntu-98-R95546020-1.pdf: 4640267 bytes, checksum: 3f5971bfd6777ebbdcc6497f41f59572 (MD5) Previous issue date: 2009 | en |
dc.description.tableofcontents | Acknowledgement i
中文摘要 ii Abstract iii 目錄 iv 名詞翻譯 x 1. 簡介 1 1.1. 背景和文獻回顧 1 1.2. 研究目的及論文架構 7 2. 逐段二次迴歸樹(Piecewise Quadratic Regression Tree) 8 2.1. 連續型變數的選擇方法 8 2.2. 二次效力變數的選擇法 10 2.3. Gram-Schmidt process 14 2.4. 樹狀結構(Tree Structure) 20 2.5. 迴歸模型的估計 21 2.6. 連動效力檢定(Interaction Effect Test) 22 2.7. 組合變數(Attribute Combination)的選擇法 24 2.8. 重新建構樹狀結構 26 2.9. 系統流程 28 3. 個案研究 30 3.1. 模擬個案 30 3.1.1. 個案一:具有連續性效力的變數 30 3.1.2. 個案二:具有連續變動效力的變數 32 3.1.3. 個案三:具有混合效力的變數 34 3.1.4. 個案四:具有逐段線性效力的變數 36 3.1.5. 個案五:組合變數 39 3.1.6. 個案六:線性效力的連續型變數 40 3.1.7. 個案七:二次效力的連續型變數 42 3.1.8. 個案八:片段且具有二次效力的連續型變數 45 3.1.9. 個案九:多個片段且具有二次效力的連續型變數 48 3.1.10. 個案十:拋物線(Parabolic Curve) 51 3.1.11. 個案十一:兩個拋物線 53 3.1.12. 個案十二:三個拋物線 56 3.1.13. 個案十三:線性和拋物線結合的資料 59 3.1.14. 個案十四:”V”型的資料 62 3.1.15. 個案十五:”W”型的資料 65 3.1.16. 個案十六:波浪狀效力的資料 68 3.1.17. 個案十七:連續波浪狀效力的資料 71 3.2. 實際個案:半導體製程的真實例子 74 4. 結論 78 Reference 79 | |
dc.language.iso | zh-TW | |
dc.title | 提高樣本使用效率之逐段二次迴歸樹 | zh_TW |
dc.title | Sample-Efficient Piecewise Quadratic Regression Tree | en |
dc.type | Thesis | |
dc.date.schoolyear | 97-1 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 范治民,楊烽正 | |
dc.subject.keyword | 二次效力,迴歸樹,多重迴歸,共線性,Gram-Schmidt, | zh_TW |
dc.subject.keyword | quadratic effects,regression tree,multiple regression,Gram-Schmidt Process,collinearity, | en |
dc.relation.page | 79 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2009-02-13 | |
dc.contributor.author-college | 工學院 | zh_TW |
dc.contributor.author-dept | 工業工程學研究所 | zh_TW |
顯示於系所單位: | 工業工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-98-1.pdf 目前未授權公開取用 | 4.53 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。