"Group sequential procedures based on Ranked Set Sampling, sequential change-point and shrinkage estimation in correlated data"

在今天的地球村中，复杂的数据可以很容易地实时获得。这就要求找到有效的统计方法来估计和监测控制数据的参数。在这个项目中，我提出了在统计估计和测试过程中纳入先验非样本信息（如专家意见）的有效方法。这些方法将改进我们处理来自临床试验、纵向、空间和集群相关数据的数据分析的方法。我的建议的第二个目标是建立基于表现时间相关性的结果的参数监测的统计方法。这些方法可用于传染病的实时监测、卫生保健绩效监测和金融市场监测。

In today's golobal village, complex data has become easily available in real-time. This has led to a higher demand in finding efficient statistical methods for estimation and surveillance of parameters governing the data. In this project, I propose efficient methods which incorporate prior nonsample information (such as expert opinion) in the statistical estimation and testing procedures. These methodlogies will improve the ways we tackle analysis of data from clinical trials, longitudinal, spatial and cluster correlated data. A second objective of my proposal is to build statistical methods for the surveillance of parameters based on outcomes that present temporal correlations. These methods have applications in real-time surveillance of infectious diseases, health care performace monitoring and in monitoring financial markets.