Spatial Inference with Application to Astronomy

空间推理及其在天文学中的应用

• 批准号: 0507687
• 负责人: Ji Meng Loh
• 金额: $ 15.57万
• 依托单位: Columbia University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-07-01 至 2010-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0507687&HistoricalAwards=false
• 关键词: Spatial; Inference; Application; Astronomy

AST-0507687LohAstronomers study large-scale structure, usually with second-order statistics of clustering, in order to understand the Universe and its evolution. To get error estimates, bootstrap is often used. This project will extend recently developed methods that dramatically reduce standard errors by considering the dependence of absorbers across different radial lines of observation, and will apply these techniques to a new larger dataset. The research work will also compare different measures of second-order structure in spatial point processes, and study methods of resampling spatial data. Extending the method to variable size absorbers, chemical element differences, and third-order characteristics, should provide clustering estimates of unprecedented precision. Analysis of surveys and simulations using various estimators should improve understanding of the procedures, and provide recommendations of appropriate measures for different situations. Resampling spatial processes is complicated and computationally intensive, and a new faster method will be tested on real data and simulations using both simple regions and more realistic, more complex regions. This work will advance statistical understanding of both theoretical and applied issues involved in estimating clustering and resampling spatial data.All of these results will be important for applied statistical work in many fields, and will be disseminated as widely as possible to the statistics and astronomy communities. The research also involves significant interdisciplinary collaboration and training, including members of under-represented groups.

AST-0507687 Loh天文学家研究大尺度结构，通常使用聚类的二阶统计，以了解宇宙及其演化。 为了得到误差估计，通常使用Bootstrap。 该项目将扩展最近开发的方法，通过考虑吸收体在不同径向观测线上的依赖性，大大降低标准误差，并将这些技术应用于一个新的更大的数据集。 研究工作还将比较空间点过程中二阶结构的不同度量，并研究空间数据的恢复方法。 将该方法扩展到可变尺寸吸收体、化学元素差异和三阶特性，应提供前所未有的精度的聚类估计。 使用各种估算方法对调查和模拟进行分析，应能增进对程序的理解，并就不同情况下的适当措施提出建议。 恢复空间过程是复杂的和计算密集型的，将使用简单区域和更真实、更复杂的区域在真实的数据和模拟上测试一种新的更快的方法。 这项工作将促进对估计聚类和重新排列空间数据所涉理论和应用问题的统计理解，所有这些结果对许多领域的应用统计工作都很重要，并将尽可能广泛地向统计和天文学界传播。 研究还涉及重要的跨学科合作和培训，包括代表性不足的群体的成员。