Full metadata record
DC FieldValueLanguage
dc.contributor.author林泰佐en_US
dc.contributor.authorTai-Tso Linen_US
dc.contributor.author王維菁en_US
dc.contributor.authorWeijing Wangen_US
dc.date.accessioned2014-12-12T01:17:23Z-
dc.date.available2014-12-12T01:17:23Z-
dc.date.issued2007en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#GT009526517en_US
dc.identifier.urihttp://hdl.handle.net/11536/38996-
dc.description.abstract線性轉換模型是相當彈性的半母數迴歸模型。倖存分析常見的Cox模型與Odds模型,皆是線性轉換模型的特例。近年來許多研究針對線性轉換模型提出半母數推論方法,一套分析方法卻有廣泛的應用價值,是其吸引人的地方。我們以古典推論理論的兩個原則(動差法和概式法)為架構,檢視現有文獻的建構方式,希望此統整的角度有助辨識不同方法的特質。此外針對設限資料,我們除了討論現有文獻的做法外,並提出一個新的方法。所有的方法均透過模擬實驗檢驗其表現。zh_TW
dc.description.abstractThe linear transformation model, which includes the proportional hazard model and the proportional odds model, has received considerable attentions in recent years due to its flexibility. In the thesis, we consider semi-parametric estimation for the regression parameter. We review existing literature under the framework of classical inference theory. Specifically we will see how these “old” principles, namely method of moment and likelihood estimation, are applied to the modern estimation problem which involves an infinite dimensional nuisance parameter in the model formulation. After examining common techniques of handling censored data, we also propose a new approach. All the methods are evaluated by Monte Carlo simulations.en_US
dc.language.isoen_USen_US
dc.subjectCox PH 模型zh_TW
dc.subjectOdds 模型zh_TW
dc.subject動差法zh_TW
dc.subjectCox proportional hazard modelen_US
dc.subjectProportional odds modelen_US
dc.subjectCounting processen_US
dc.subjectInverse probability weightingen_US
dc.subjectMethod of momenten_US
dc.subjectProfile likelihooden_US
dc.title線性轉換模型之統計推論zh_TW
dc.titleStatistical Inference for Linear Transformation Modelsen_US
dc.typeThesisen_US
dc.contributor.department統計學研究所zh_TW
Appears in Collections:Thesis


Files in This Item:

  1. 651701.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.