簡易檢索 / 詳目顯示

研究生: 劉季衡
Liu, Chi-Heng
論文名稱: 比較在不同相關性結構的結果依賴抽樣下多變量存活資料的統計方法
Statistical methods for comparison of multivariate survival data under outcome-dependent sampling with different working correlation structures
指導教授: 呂翠珊
Lu, Tsui-Shan
口試委員: 徐雅甄
Hsu, Ya-Chen
張少同
Chang, Shao-Tung
呂翠珊
Lu, Tsui-Shan
口試日期: 2022/06/29
學位類別: 碩士
Master
系所名稱: 數學系
Department of Mathematics
論文出版年: 2022
畢業學年度: 110
語文別: 英文
論文頁數: 55
中文關鍵詞: 結果依賴抽樣設計相關性結構加速失效模型
英文關鍵詞: outcome-dependent sampling design, correlation structure, accelerated failure time model
研究方法: 紮根理論法比較研究
DOI URL: http://doi.org/10.6345/NTNU202201108
論文種類: 學術論文
相關次數: 點閱:96下載:10
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 結果依賴抽樣設計 (outcome-dependent sampling design, ODS) 是一種已被證實可以提高有效性且降低研究成本的抽樣方法。其利用簡單隨機抽樣以及機率依賴抽樣來取得研究樣本。而現今 ODS 已被發展且調整至多維度資料當中。在最近的文獻中,ODS 已被延伸至抽取擁有多個觀測值的樣本,譬如來自同一病人的多重疾病或是多個家族成員中的某種疾病。然而到目前為止,只有每個集群包含兩個觀測值的情況被考慮過。
    在此研究中,我們對於在加速失效模型 (accelerated failure time, AFT) 下將 ODS抽樣方法推廣至更高維度的存活資料中且考慮觀測值之間的相關性結構感到興趣。我們通過進行廣泛的模擬實驗來對比多種抽樣設計以及不同的相關性結構,其中包含當補充樣本數較少的處境。此外,我們還建立最佳化的設計和配置來充分提高有效性。最後我們套用所提出的方法於真實牙齒資料中進行分析。

    The outcome-dependent sampling design (ODS) is a sampling design that has been shown to improve efficiency and reduce study cost. The ODS sample consists of not only a simple random sample (SRS) but also some supplemental samples through a probability sampling process which depends on the outcome level. Today, it has been developed to adjust for multivariate data under ODS. In the recent literatures, the multivariate ODS design has been extended to select the samples with multiple observations such as multiple disease outcomes from one patient or any certain disease outcome from family members. However, only the case where each cluster has two observations was discussed so far.
    In this thesis, we are interested in extending multivariate outcome-dependent sampling design to higher dimensional failure time data under an accelerated failure time (AFT) model, and consider the correlation structures between the observations. We compare several sampling designs along with different correlation structures by conducting extensive simulation studies, including the situation particularly when the sizes of the supplemental samples are small. In addition, we establish optimal designs and allocations, which can substantially gain more efficiency. We apply our proposed methods to the dental data for analysis in the end.

    Chapter 1 Introduction 1 Chapter 2 Model Design and Estimation 3 Section 2.1 AFT Model 3 Section 2.2 Multivariate ODS designs 3 Section 2.3 Estimation of model parameters 6 Chapter 3 Simulation Studies 8 Section 3.1 Data generation 8 Section 3.2 Results 9 Section 3.3 The optimal design 41 Chapter 4 Analysis of the Dental Study 46 Section 4.1 Background 46 Section 4.2 Data Processing and Model Fitting 46 Section 4.3 Results 47 Chapter 5 Conclusions and Discussions 53 . . References 54

    [1] Buckley, J., & James, I. (1979). Linear regression with censored data. Biometrika, 66, 429-436.
    [2] Caplan, D. J., Li, Y., Wang, W., Kang, S., Marchini, L., Cowen, H. J., & Yan, J. (2019). Dental Restoration Longevity among Geriatric and Special Needs Patients. JDR Clinical and Translational Research, 4, 41-48.
    [3] Chiou, S. H., Kang, S., Kim, J. & Yan, J. (2014a). Marginal semiparametric multivariate accelerated failure time model with generalized estimating equations. Lifetime Data Analysis, 20, 599-618.
    [4] Chiou, S. H., Kang, S., & Yan, J. (2014b). Fitting accelerated failure time model in routine survival analysis with R package aftgee. Journal of Statistical Software, 61, 1-23.
    [5] Ding, J., Lu, T. S., Cai, J., & Zhou, H. (2017). Recent progresses in outcome-dependent sampling with failure time data. Lifetime Data Analysis, 23, 57-82.
    [6] Higgins, M., Province, M., Heiss, G., Eckfeldt, J., Ellison, R. C., Folsom, A. R., Rao, D. C., Sprafka, J. M., & Williams, R. (1996). NHLBI Family Heart Study: objectives and design. American Journal of Epidemiology, 143, 1219-1228.
    [7] Jin, Z., Lin, D. Y., & Ying, Z. (2006). On least-squares regression with censored data. Biometrika, 93, 147-161.
    [8] Johnson, L. M., & Strawderman, R. L. (2009). Induced smoothing for the semiparametric accelerated failure time model: asymptotics and extensions to clustered data. Biometrika, 96, 577-590.
    [9] Liang, K. Y., & Zeger, S. L. (1986). Longitudinal data analysis using generalized linear models. Biometrika, 73, 13-22.
    [10] Lu, T. S., Kang, S., & Zhou, H. (2020). Semiparametric accelerated failure time modeling for multivariate failure times under multivariate outcome-dependent sampling designs. Statistics and Its Interface, 13, 373-383.
    [11] Lu, T. S., Longnecker, M. P., & Zhou, H. (2018). Statistical inferences for data from studies conducted with an aggregated multivariate outcome-dependent sample design. Statistics in Medicine, 36, 985-997.
    [12] Zhou, H., Weaver, M. A., Qin, J., Longnecker, M. P., & Wang, M. C. (2002). A semi-parametric empirical likelihood method for data from an outcome-dependent sampling scheme with a continuous outcome. Biometrics, 58, 413-421.

    下載圖示
    QR CODE