透過您的圖書館登入
IP:3.141.30.162
  • 學位論文

區間限制資料下勝算之研究

Estimation of Odds for Interval-Censored Data

指導教授 : 吳裕振

摘要


一個疾病的勝算是大家想知道的,例如得到流感的人,最後死亡人數和康復人數的比值是多少?因此,本論文主要研究對型二區間設限資料的勝算做估計,我們使用伯氏多項式來描述勝算統計方法,用最大概似估計的方法,但要計算最大概似估計非常複雜且困難。因此使用了馬可夫鏈蒙地卡羅法─Simulated Annealing來估計其參數,並且做了統計模擬,其結果表現相當不錯。

並列摘要


We all want to know the odds of a disease. For instance, what's the ratio of the last death toll and the number of rehabilitation for people who get the flu? The thesis aims at studying the odds of Case Ⅱ interval-censored data. We apply the Bernstein polynomial to describe the statistical method of odds. We use the maximum approximate estimation method, however ; calculating the maximum approximate estimate is very complicated and difficult. Therefore, we adopt Markov chain Monte Carlo - Simulated Annealing to estimate its parameters, and we also conducted a lot of statistical simulation, and the results are quite satisfied.

參考文獻


[5 ]楊雅雯 (2009) 使用伯氏多項式對存活現狀數據之最大概似估計,
[6 ]賴京鈺 (2011) 存活現狀數據下風險函數之最大概似估計,中原大
[8 ]丁怡方 (2015) 存活右設限和現狀數據混合資料之分佈函數估計,中
[9 ]許時淮 (2015) 現狀數據資料下勝算比之研究,中原大學碩士論文
[2 ]Christian P. Robert, George Casella (1999), “Monte Carlo Statistical Methods ” , Springer-Verlag, New York

延伸閱讀