首页> 外文期刊>Statistics and computing >Functional variance estimation using penalized splines with principal component analysis
【24h】

Functional variance estimation using penalized splines with principal component analysis

机译:使用惩罚样条和主成分分析的功能方差估计

获取原文
获取原文并翻译 | 示例
           

摘要

In many fields of empirical research one is faced with observations arising from a functional process. If so, classical multivariate methods are often not feasible or appropriate to explore the data at hand and functional data analysis is prevailing. In this paper we present a method for joint modeling of mean and variance in longitudinal data using penalized splines. Unlike previous approaches we model both components simultaneously via rich spline bases. Estimation as well as smoothing parameter selection is carried out using a mixed model framework. The resulting smooth covariance structures are then used to perform principal component analysis. We illustrate our approach by several simulations and an application to financial interest data.
机译:在许多实证研究领域,人们都面临着功能过程产生的观察结果。如果是这样,经典的多变量方法通常不可行或不适合探索手头的数据,并且功能数据分析很盛行。在本文中,我们提出了一种使用罚样条对纵向数据的均值和方差进行联合建模的方法。与以前的方法不同,我们通过丰富的样条库同时对两个组件进行建模。使用混合模型框架进行估计以及平滑参数选择。然后将所得的平滑协方差结构用于执行主成分分析。我们通过一些模拟以及对财务权益数据的应用来说明我们的方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号