透過您的圖書館登入
IP:18.118.0.240
  • 學位論文

相關右設限資料下存活函數之估計

The estimation of survival function under dependent right censored data

指導教授 : 謝進見
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


在存活分析中我們常假設獨立設限的情況下利用Kaplan-Meier 的方法估計存活函數。 但在相依的情況下如何估計存活函數呢? 在此篇文章中我們將探討在相關右設限資料下, 給定存活時間和設限時間之相關性,利用向右重新分配演算法對有興趣的存活時間之存活函數進行估計, 並且利用Archimedean copula 模型處理存活時間與設限時間的相關性。 根據模擬的結果顯示我們所提出的向右重新分配演算法表現得還不錯。 最後,我們利用此方法對一筆卵巢癌的資料進行分析。

並列摘要


In the survival analysis, we usually assume the independent censored relationship to estimate the survival function with Kaplan-Meier method. In this thesis, we investigate how to estimate the survival function under the situation of dependent censored. Here, we proposed the redistribute-to-the-right algorithm with a given association between the failure time T and censoring time C to estimate the survival function of the failure time under the dependent right censored data. We use the Archimedean copula (AC) model assumptions to specify the dependence between T and C. Simulation studies con rm the redistribute-to-the-right algorithm performs well. Finally, we use the redistribute-to-the-right algorithm to analyze the ovarian cancer data.

參考文獻


Right Algorithm: Application to the Analysis of Censored Cost
[2] Efron, B. (1967). The two sample problem with censored data. Proc.
[3] Efron, B. and Stein, C. (1981). The Jackknife Estimate of Variance.
vol. 40, Issue 4, pp. 280-283.
[5] Genest, C. (1987). Frank's family of bivariate distributions. Oxford

延伸閱讀