误差为NQD序列的删失部分线性模型的强相合性研究.pdfVIP

南京师范大学 硕士学位论文 误差为NQD序列的删失部分线性模型的强相合性 姓名：薛欣 申请学位级别：硕士 专业：数学；运筹学与控制论 指导教师：刘国祥 2009-04-26 , , , , , . , [8] , , . , . , . . , , , , , , . , , , [8]. , : (1); (2) . : ; ; ; ii Abstract Survival analysis is one of the current popular topics in statistics, which is widely used in all kinds of fields such as medicine, life science, reliability engineering and so on, and mainly focus on the methods and its theory of statistical inference of censored data, an important aspect of which is regression analysis with censored data. Classical censored regression is mostly based on the independent errors, for example, Wang [8] considered the partial linear model where is randomly right-censored variable with known and unknown distribution, defined the estimates of and , and proved their strong consistency. However, the assumption of independence is too strict and there is always some kind of dependent structure among errors. In this paper, under the assumption of known censored variable distribution, we continue to consider the strong consistency of the partial censored linear model with errors. This article consists of the following three chapters. In Chapter 1, we introduce the survival analysis and censored data, partial linear censored model and