Finite mixture models and their applications

有限混合模型及其应用

• 批准号: 371502-2009
• 负责人: Li, Pengfei
• 金额: $ 1.53万
• 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: 2013
• 资助国家: 加拿大
• 起止时间: 2013-01-01 至 2014-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=528235
• 关键词: Finite; mixture; models; their; applications

Finite mixture models are widely applied to data sets in which several homogeneous populations are mixed together. In financial economics, mixtures of three normal distributions are often applied to daily stock returns due to the existence of three risk regimes: low, normal and high. The number of homogeneous populations in a finite mixture is of great interest in applications. For example in statistical genetics, this number reflects the number of major genes in action. In the first project, a novel and practically convenient method, called the EM-test, will be developed for testing the number of homogeneous populations in a finite mixture.

有限混合模型广泛应用于多个同质总体混合的数据集。在金融经济学中，由于存在三种风险状态：低，正常和高，三种正态分布的混合物经常被应用于每日股票收益率。有限混合体中齐次总体的个数是一个非常有意义的问题。例如，在统计遗传学中，这个数字反映了起作用的主要基因的数量。 在第一个项目中，一个新的和实际上方便的方法，称为EM测试，将开发用于测试在有限的混合均匀的人口的数量。