Geometric methods in statistics

统计学中的几何方法

• 批准号: 311995-2008
• 负责人: Marriott, Paul
• 金额: $ 1.6万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2010
• 资助国家: 加拿大
• 起止时间: 2010-01-01 至 2011-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=457240
• 关键词: Geometric; methods; statistics

A great deal of work that is done by statisticians exploits the idea of a statistical model. Such a model enables analysts to, amongst other things, quantify their uncertainty about the numerical value of a real world phenomenon, make predictions of events in the future and to test scientific hypotheses. The model itself though is just a tool and like any other needs to be selected with care. The objective of the research is to find ways of evaluating the effect of the choice of a particular model in a given problem. Somewhat surprisingly it turns out that tools from the world of geometry are very useful in this assessment. The project will develop these geometrically based tools in ways that the working statistician will be able to exploit.

统计学家所做的大量工作都利用了统计模型的概念。这种模型使分析师能够，除其他外，量化他们对真实的世界现象的数值的不确定性，预测未来的事件，并测试科学假设。 模型本身只是一个工具，像任何其他需要小心选择。研究的目的是找到评估在给定问题中选择特定模型的效果的方法。令人惊讶的是，几何世界的工具在这种评估中非常有用。该项目将开发这些几何工具的方式，工作统计人员将能够利用。