對型Ⅱ區間設限資料有交互作用共變量的貝氏柯斯迴歸模型之研究

Research on the Bayesian Cox-Regression Model with Interaction Covariate for Type Ⅱ Interval-Censored Data

曾琬萱(WAN-XUAN TSENG)

指導教授 ： 吳裕振

中原大學/理學院/應用數學系/碩士(2022年)

https://doi.org/10.6840/cycu202201060

摘要

若患有該疾病的風險因子有n個，我們想知道此n個風險因子是否互相具有交互作用，在生物醫學統計上是非常重要的。而本論文只探討兩個風險因子下，並且把風險因子架構在柯斯迴歸模型上，用伯氏多項式描述其基線累積風險函數，對於研究型II區間設限存活資料，進行貝氏統計來分析兩個風險因子是否會產生交互作用。在參數貝氏估計用馬可夫鏈蒙地卡羅來計算，其模擬結果上具有大樣本的性質，表現相當良好。

關鍵字

貝氏柯斯迴歸 ； 伯氏多項式 ； 交互作用 ； 型II區間設限資料

並列摘要

If there are n risk factors for the disease, we want to know whether the n risk factors interact with each other, which is very important in biomedical statistics.However, this paper only discusses two risk factors, constructing the risk factors on the Cox regression model and using the Bernstein polynomial to describe its baseline cumulative risk function. For the research-type II interval-limited survival data, Bayesian statistics are used to analyze whether or not two risk factors interact. In the parameter Bayesian estimation, the Markov chain Monte Carlo is used for calculation purposes, and their performance is considered quite practical since their simulation results have the property of large samples.

並列關鍵字

Bayesian cox regression ； Bernstein polynomial ； Interaction ； Type Ⅱ Interval-Censored Data

參考文獻

1. Baojiang Chen, Jing Qin, Ao Yuan(2021), Using the accelerated failure time model to analyze current status data with misclassified covariates. Electronic Journal of Statistics, Vol.15, 1372–1394.

2. Chang,I.S., Wen,C,C and Wu,Y.J(2007),”Bayesian Survival Analysis for Cox-model Using Bernstein polynomials”. (初稿)

3. Chang I.S., Hsiung C.A , Wu Y.J., Yang C.C. (2005),” Bayesian Survival Analysis Using Bernstein polynomials”,Seandinanian Journal of Statistics Vol.32 P.447-466.

4. Yu Jichang, Zhou Haibo, Cai Jianwen (2021), Accelerated failure time model for data from outcome-dependent sampling. Lifetime Data Anal.,27(1),p.15~37.

5. Robert, C.P. & Casella, G.(1999). Monte Carlo statistical methods. Springer-Verlag, New York.