標題: | 非混合治癒模型之統計推論 Statistical Inference for Cure Models-The Non-mixture Approach |
作者: | 陳瀅竹 Chen, Ying-jhu 王維菁 Wang, Weijing 統計學研究所 |
關鍵字: | 治癒率;非混合治癒模型;模型診斷;Cure rate;Non-mixture model;Model checking |
公開日期: | 2008 |
摘要: | 在存活分析中,若在實驗期間能無限延長下,有部分的資料能夠永久存活或者不會經歷到所感興趣的事件發生,必須考慮使用治癒模型。本論文主要探討的是Tsodikov (1998) 所提出的非混合治癒模型,討論在數學模型的架構下代表的生物意義,藉模擬討論在有母數與無母數的統計推論方法中對治癒率以及發病時間函數參數估計的影響。最後在有解釋變數對模型的影響中,提供符合比例風險迴歸模型(PH) 與加速失敗時間模式(accelerated failure time model)的圖型診斷方法。 Cure models are survival models which allows for the existence of cure despite of long-term follow-up. In this thesis, we consider statistic inference for cure models based on the non-mixture approach proposed by Tsodikov(1998). This model has some interesting biological interpretations. Parametric and semi-parametric methods will be investigated. We propose a model diagnostic tool to verify the form of regression models. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT079626523 http://hdl.handle.net/11536/42684 |
Appears in Collections: | Thesis |
Files in This Item:
If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.