學術產出-Theses

Article View/Open

Publication Export

Google ScholarTM

政大圖書館

Citation Infomation

  • No doi shows Citation Infomation
題名 雙向分類中有影響力觀察值的診斷
作者 葉昭瑛
YE, ZHAO-YING
貢獻者 宋傳欽
SONG, CHUAN-QIN
葉昭瑛
YE, ZHAO-YING
關鍵詞 雙向分類
影響力
觀察值
日期 1992
1991
上傳時間 2-May-2016 17:07:04 (UTC+8)
摘要 由實驗所得之雙向分類的反應值,常因人為記錄的疏忽,或是儀器設備的測量誤差,也可能是某些潛在因素,帶來了一些有問題的觀察值,這些問題點的存在,可能會與其他資料點有明顯的距離差異或位置偏離,同時亦可能會影響模型之參數估計,導致研究者選擇了不當的模型,為了避免此缺憾,本文乃針對雙向分類之反應值被刪去一個或二個的情況下,導出AP及Cook統計量的公式,在進行模式選擇之前,利用這二個統計量,先對所蒐集到的資料做一番診斷,以便找出離群值或是對參數估計有影響力的觀察值。當然,本文診斷雙向分類中有影響力觀察值的探測工作是屬於先驅工作中的一小部分,其目的不在於確認有那些不尋常的點,而是在懷擬那些點可能會有問題,以提供往後分析工作更多的資訊。
參考文獻 【l】. Cook, R. D. (1977), "Detection of Influential Observations in Regression
     ," Technometrics, 19, 15-18.
     【2】. Andrews, D. F., and Pregibon , D. (1978), "Finding the Outliers
     That Matter ," Journal of the Royal Statistical Society, Ser. B, 40,85-93.
     【3】. Hoaglin, D. C., and Welsch, R. E. (1978), "The Hat matrix in
     Regression and Anova," The A`merican Statistician, 32, 17-22.
     【4】. Draper, N. R., and John, J. A. (1981)," Influential Observations
     and Outliers in Regression," Technometrics, 23, 21-26.
     【S】. John, J. A., and Draper, N. R. (1978),"On Testing for Two Outliers
     or One Outlier in Two -Way Tables," Technometrics, 20, 69-78.
     【6】. Gentleman, J. F., and Wille, M. B. (1975),"Detecting Outliers II
     :Supplementing the Direct Analysis of Residuals," Binometrics, 31,387-410.
     【7】. Beckman, R. J. and Trussell, H. J. (1974)," The distribution of
     an arbitrary studentized residual and the effects of updating in
     multiple regression,J. Amer. Statist. Assoc., 69, 199-20l.
     【8】. Schall, R., and Dunne, T. T. (1990),"Influential Variables in Linear
     Regression," Technometrics, 32, 323-330.
     【9】. Seber, G.A.F. (1977), Linear Regression Analysis, John Wiley,New York.
     【10】 Draper, N. R. and Smith, H. (1981),Applied Regression Analysis
     (2nd, ed.), John Wiley, New York.
     【11】. Anderson, R. L., and Bancroft, T. A.(1952),Statistical Theory in
     Research McGraw-Hill Book Co., New York.
描述 碩士
國立政治大學
應用數學系
資料來源 http://thesis.lib.nccu.edu.tw/record/#B2002004728
資料類型 thesis
dc.contributor.advisor 宋傳欽zh_TW
dc.contributor.advisor SONG, CHUAN-QINen_US
dc.contributor.author (Authors) 葉昭瑛zh_TW
dc.contributor.author (Authors) YE, ZHAO-YINGen_US
dc.creator (作者) 葉昭瑛zh_TW
dc.creator (作者) YE, ZHAO-YINGen_US
dc.date (日期) 1992en_US
dc.date (日期) 1991en_US
dc.date.accessioned 2-May-2016 17:07:04 (UTC+8)-
dc.date.available 2-May-2016 17:07:04 (UTC+8)-
dc.date.issued (上傳時間) 2-May-2016 17:07:04 (UTC+8)-
dc.identifier (Other Identifiers) B2002004728en_US
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/89752-
dc.description (描述) 碩士zh_TW
dc.description (描述) 國立政治大學zh_TW
dc.description (描述) 應用數學系zh_TW
dc.description.abstract (摘要) 由實驗所得之雙向分類的反應值,常因人為記錄的疏忽,或是儀器設備的測量誤差,也可能是某些潛在因素,帶來了一些有問題的觀察值,這些問題點的存在,可能會與其他資料點有明顯的距離差異或位置偏離,同時亦可能會影響模型之參數估計,導致研究者選擇了不當的模型,為了避免此缺憾,本文乃針對雙向分類之反應值被刪去一個或二個的情況下,導出AP及Cook統計量的公式,在進行模式選擇之前,利用這二個統計量,先對所蒐集到的資料做一番診斷,以便找出離群值或是對參數估計有影響力的觀察值。當然,本文診斷雙向分類中有影響力觀察值的探測工作是屬於先驅工作中的一小部分,其目的不在於確認有那些不尋常的點,而是在懷擬那些點可能會有問題,以提供往後分析工作更多的資訊。zh_TW
dc.description.tableofcontents 第一章緒論..........1
     第一節:前言..........1
     第二節:本文架構..........2
     第二章線性迴歸模型中有關觀察值診斷之文獻回顧..........3
     第一節:帽子矩陣所扮演的角色..........3
     第二節:Qk統計量的介紹..........5
     第三節:AP統計量的介紹..........6
     第四節:Cook統計量的介紹..........8
     第三章雙向分類中,刪去一個或兩個觀察值時AP及Cook統計量的推導..........12
     第一節:前言..........12
     第二節:帽子矩陣的導..........14
     第三節:Q2統計量的推導..........17
     第四節:AP統計量的推導..........21
     第五節:Cook統計量的推導..........24
     第六節:討論..........26
     第四章實例應用與分析..........37
     第一節:資料描述..........37
     第二節:結果分析與討論..........38
     第三節:結論..........43
     附錄..........45
     參考文獻..........49
zh_TW
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#B2002004728en_US
dc.subject (關鍵詞) 雙向分類zh_TW
dc.subject (關鍵詞) 影響力zh_TW
dc.subject (關鍵詞) 觀察值zh_TW
dc.title (題名) 雙向分類中有影響力觀察值的診斷zh_TW
dc.type (資料類型) thesisen_US
dc.relation.reference (參考文獻) 【l】. Cook, R. D. (1977), "Detection of Influential Observations in Regression
     ," Technometrics, 19, 15-18.
     【2】. Andrews, D. F., and Pregibon , D. (1978), "Finding the Outliers
     That Matter ," Journal of the Royal Statistical Society, Ser. B, 40,85-93.
     【3】. Hoaglin, D. C., and Welsch, R. E. (1978), "The Hat matrix in
     Regression and Anova," The A`merican Statistician, 32, 17-22.
     【4】. Draper, N. R., and John, J. A. (1981)," Influential Observations
     and Outliers in Regression," Technometrics, 23, 21-26.
     【S】. John, J. A., and Draper, N. R. (1978),"On Testing for Two Outliers
     or One Outlier in Two -Way Tables," Technometrics, 20, 69-78.
     【6】. Gentleman, J. F., and Wille, M. B. (1975),"Detecting Outliers II
     :Supplementing the Direct Analysis of Residuals," Binometrics, 31,387-410.
     【7】. Beckman, R. J. and Trussell, H. J. (1974)," The distribution of
     an arbitrary studentized residual and the effects of updating in
     multiple regression,J. Amer. Statist. Assoc., 69, 199-20l.
     【8】. Schall, R., and Dunne, T. T. (1990),"Influential Variables in Linear
     Regression," Technometrics, 32, 323-330.
     【9】. Seber, G.A.F. (1977), Linear Regression Analysis, John Wiley,New York.
     【10】 Draper, N. R. and Smith, H. (1981),Applied Regression Analysis
     (2nd, ed.), John Wiley, New York.
     【11】. Anderson, R. L., and Bancroft, T. A.(1952),Statistical Theory in
     Research McGraw-Hill Book Co., New York.
zh_TW