Inference strategies with applications and boostrapping

具有应用程序和 boostrapping 的推理策略

• 批准号: 98832-2006
• 负责人: Ahmed, Syed
• 金额: $ 0.87万
• 依托单位: University of Windsor
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2007
• 资助国家: 加拿大
• 起止时间: 2007-01-01 至 2008-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=366145
• 关键词: Inference; strategies; applications; boostrapping

This research is primarily concerned with the development of optimal tools to help in the analysis of complex data.  For example, how can DNA microarray data from various sources be combined, showing the promise of immense possible gains in statistical inferential accuracy?  How can uncertain prior information be incorporated into the current estimation process?  How can uncertainty concerning the appropriate statistical model-estimator to use in representing the data sampling process be dealt with?    These questions, and many more, can be better understood through appropriate statistical inferential tools.   The proposed research  demonstrates well-defined data-basesd estimation techniques. The focus of this research program is on the development and application of statistical inference methodologies which performs better than the existing methods. Statistical inference is the process of reasoning from observed data back to its underlying mechanism. Shrinkage and empirical Bayes methods provide useful techniques for combining data from various sources.  B.  Efron (Newsletter of the Royal Statistical Society January, 1995) predicted that shrinkage and empirical Bayes methodology would be a major area of statistical research for the early 21st century.    More recently, several authors considered a new approach to likelihood that weigh components differentially.  Interestingly, this leads to shrinkage type estimators that also feature prominently in empirical Bayes methodology.   New scientific technology, exemplified by DNA microarrays, has suddenly revived interest in these methods. These techniques will advance knowledge because of their applicability to a larger class of problems thereby extending our ability to solve advanced statistical inference problems in health-care, environmetrics, engineering and social sciences.

本研究主要关注于开发最优工具来帮助分析复杂数据。例如，如何将来自不同来源的DNA微阵列数据结合起来，以显示在统计推断准确性方面可能获得的巨大收益？如何将不确定的先验信息纳入当前的估计过程？如何处理与用于表示数据抽样过程的适当统计模型估计器有关的不确定性？这些问题，以及更多的问题，可以通过适当的统计推断工具来更好地理解。提出的研究展示了定义良好的基于数据的估计技术。本研究计划的重点是统计推理方法的开发和应用，该方法比现有方法性能更好。统计推理是从观察到的数据推理到其潜在机制的过程。收缩和经验贝叶斯方法为组合来自不同来源的数据提供了有用的技术。B. Efron（1995年1月皇家统计学会通讯）预测，萎缩和经验贝叶斯方法将是21世纪初统计研究的主要领域。最近，几位作者考虑了一种新的可能性方法，即对成分进行不同的加权。有趣的是，这导致收缩型估计器在经验贝叶斯方法中也很突出。以DNA微阵列为代表的新科学技术，突然重新唤起了人们对这些方法的兴趣。这些技术将促进知识的发展，因为它们适用于更大范围的问题，从而扩展我们解决医疗保健、环境计量、工程和社会科学领域的高级统计推断问题的能力。