Mathematical Sciences: Statistical Intervals

数学科学：统计区间

• 批准号: 9307499
• 负责人: J.T. Hwang
• 金额: --
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Continuing grant
• 财政年份: 1993
• 资助国家: 美国
• 起止时间: 1993-08-01 至 1997-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9307499&HistoricalAwards=false
• 关键词: Mathematical; Sciences; Statistical; Intervals

Hwang will work on two areas in statistics: a theory concerning the conditional analysis and a theory about prediction based on a measurement error model. In the first part of the proposal, the principal investigator proposes to examine the estimated confidence approach by applying it to the Fieller's problem and by relating it to Fisher's exact tests for contingency tables. The second part of the research will focus on construction of prediction intervals based on measurement error models, which arise in many applications. The examples that motivate the principal investigator in this research are about the prediction problems of compressive strengths of concrete based on nondestructive tests. The principal investigator will work on two areas: conditional inference and prediction. For the first area, Hwang shall determine which statistical procedures for evaluating the probability of a particular event can take advantage of lucky data better than others; in other words, which procedures work better conditionally. The event involved refers to whether a particular interval (based on data) covers some ratios of unknown parameters. In the second area, Hwang's focus will be on prediction by intervals, where the studies are more applicable to U.S. society. The aim of the research is to construct a prediction interval for the compressive strength of a wall, a bridge, or a building based on a nondestructive method such as pulse velocity of penetrating sound waves. Success of this project may help to determine, without destroying any part of the bridge, whether the bridge is weak enough that it causes concern.

黄将致力于统计学的两个领域：条件分析理论和基于测量误差模型的预测理论。在提案的第一部分中，主要研究者建议通过将估计置信方法应用于 Fieller 问题并将其与 Fisher 列联表精确检验联系起来来检查估计置信方法。 研究的第二部分将重点关注基于测量误差模型的预测区间的构建，这在许多应用中都会出现。 激发主要研究者进行这项研究的例子是关于基于无损测试的混凝土抗压强度的预测问题。 首席研究员将致力于两个领域：条件推理和预测。 对于第一个领域，Hwang 应确定哪种统计程序能够比其他程序更好地利用幸运数据来评估特定事件的概率；换句话说，哪些程序在有条件的情况下效果更好。 涉及的事件是指某个特定区间（基于数据）是否涵盖了未知参数的某些比例。 在第二个领域，Hwang 的重点将是按时间间隔进行预测，这些研究更适用于美国社会。该研究的目的是基于穿透声波的脉冲速度等无损方法构建墙壁、桥梁或建筑物抗压强度的预测区间。 该项目的成功可能有助于在不破坏桥梁任何部分的情况下确定桥梁是否脆弱到引起担忧。