透過您的圖書館登入
IP:3.137.178.133
  • 學位論文

使用伯氏多項式對右設限資料勝算比模型之最大概似估計

Maximum Likelihood Estimator of Proportional Odds Model with Right Censored Data Using Bernstein Polynomials

指導教授 : 吳裕振

摘要


本篇論文是研究右設限資料, 我們增加共變量的資訊做為研究, 其模型為比例勝 算比模型(Proportional Odds Model), 其參考文獻可詳見Pettiet(1982) 以及Bennett(1983a,b), 在存活資料的分析上, 討論共變量Z 以及事件發生時間T(failure time) 之間關係的半母迴歸方法(semiparametric regression) 已被廣泛的研究, 例如比例風險模型(proportional hazards model), 比例勝算比模型(proportional odds model), 在吳孟倩(2012) 論文中也是研究右設限的比例勝算比模型, 其無母數模型參數部份, 她用伯氏多項式來描述, 用貝氏方法進行研究, 並且用馬可夫鏈蒙地卡羅法(M.C.M.C.) 去估計其參數, 而本篇論文, 模型假設與吳孟倩(2012) 一樣, 但我們採用概似函數最大估計法來估計參數, 但其演算還 是用M.C.M.C.。而且在模擬計算中, 我們有很好的表現。

並列摘要


The thesis focused on adding covariates to the right censored data with Proportional Odds Model, refer to Pettiet(1982) and Bennett(1983a,b) for more detail. For survival analysis, the semi-parametric regression has been widely used to calculate the correlation of covariate Z and the failure time T, such as the proportional hazards model and proportional odds model. The proportional odds model was also taken by Wu (2012) to calculate the nonparametric estimators per Bernstein polynomials,Bayesian Methodology and Makov Chain Monte Carlo (M.C.M.C.). The model taken byWu is used for this thesis as well to get the nonparametric estimator by maximum likelihood estimation (M.L.E.) and M.C.M.C.. We proved the model workable with excellent results by simulations.

參考文獻


[8] 吳孟倩, 貝氏對右設限資料勝算比模型之研究, 中原大學碩士論文, 2012.
in the Proportional Odds Model” Journal of the American Statistical
[3] I.S. Chang , C.A. Hsiung , Y.J. Wu , C.C. Yang “ Bayesian Survival Analysis
[1] A. W. van der Vaart “Asymptotic Statistics” , Cambridge University Press 1998.
[2] G. Casella , R.L. Berger “ Statistical inference ” , Duxbury Press 1990.

延伸閱讀