博碩士論文 952205006 完整後設資料紀錄

DC 欄位 語言
DC.contributor統計研究所zh_TW
DC.creator黃穎慈zh_TW
DC.creatorYing-tzu Huangen_US
dc.date.accessioned2008-6-23T07:39:07Z
dc.date.available2008-6-23T07:39:07Z
dc.date.issued2008
dc.identifier.urihttp://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=952205006
dc.contributor.department統計研究所zh_TW
DC.description國立中央大學zh_TW
DC.descriptionNational Central Universityen_US
dc.description.abstract在存活分析中,半母數模型通常用 Cox 比例風險模型或加速失敗時間模型來描述變數與存活資訊間的關係,而當變數的資料型態為長期追蹤資料時,可以使用聯合模型做配適,以消除因資料遺失所造成的偏誤。在使用聯合模型的文獻中,大多數都是以 Cox 比例風險模型的風險函數作為架構,搭配混合線性模型組合成聯合模型。但是當資料不符合比例性的假設時,使用 Cox 比例風險模型並不合理,因此 Tseng et al.(2005) 提出可以使用加速失敗時間模型來替換 Cox 比例風險模型。本文引用 Tseng et al.(2005) 提出之加速失敗時間模型架構下的聯合模型,介紹如何使用 EM 演算法估計未知參數與未知的基準風險函數,以及使用拔靴法估計參數的標準差。並將模型應用在 1000 隻雌性地中海果蠅的資料上,其中以果蠅的每日產卵數作為繁殖力的指標,存活時間作為老化程度的指標,以模型探討繁殖力與老化間的關係;此外依照總產卵量大小對果蠅進行分組,用以探討總產卵量的多寡是否會影響參數的估計以及變數與存活時間之間的關係。zh_TW
dc.description.abstractIn survival analysis, Cox proportional hazard model or accelerated failure time model are used to describe the relationship between variate and survival information in semi-parametric model. When the type of the variate data becomes a longitudinal data, the joint model will be suitable. Most literature would use the Cox proportional hazard model as its basis and then combine it with the mixed linear model to form the joint model. However, it will be unreasonable to use Cox proportional hazard model when the data can’’t fit in the proportionality assumption. Consequently, Tseng et al. (2005) suggests to use the accelerated failure time model instead. This thesis will use the joint model of linera mixed dffdct model and accelerated failure time model that proposed by Tseng et al. (2005), and to introduce how to estimate the unknown parameter and baseline hazard function by using EM algoritme and estimate the standard error by bootstrap as well. Besides, this model can also be applied to the study of 1000 female medflies.As for the method, we can use the reproductiive egg-laying data as the index of fecundity and the longevity as the index of ageing to observe the relationship between the fecundity and ageing. What’’s more, group the medflies according to their total amont of edd-lying data to study if it will affect the estimation of parameter and the relationship between variate and survival time.en_US
DC.subject隨機效應zh_TW
DC.subject加速失敗時間模型zh_TW
DC.subject聯合模型zh_TW
DC.subjectRandom effecten_US
DC.subjectAccelerated failure time modelen_US
DC.subjectJoint modelen_US
DC.title以聯合模型探討地中海果蠅繁殖力與老化之關係zh_TW
dc.language.isozh-TWzh-TW
DC.titleJoint modelling fecundity and aging of medflyen_US
DC.type博碩士論文zh_TW
DC.typethesisen_US
DC.publisherNational Central Universityen_US

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明