Estimating Mixture of Gaussian Processes by Kernel Smoothing.

Estimating Mixture of Gaussian Processes by Kernel Smoothing.
复制标题

通过核平滑估计高斯过程的混合

DOI:
10.1080/07350015.2013.868084
复制
发表时间:
2014
期刊:
Journal of business & economic statistics : a publication of the American Statistical Association
影响因子:
--
通讯作者:
Yao W
Yao W
中科院分区:
其他
文献类型:
--
作者:
Huang M;Li R;Wang H;Yao W

文献摘要

参考文献

被引文献

相似文献

当函数数据不均匀时,例如当数据集中有多类函数曲线时,传统的估计方法可能会失效。在这篇文章中,我们提出了一种新的混合高斯过程的估计方法,结合了数据的函数性质和非齐次性质。我们的方法可以看作是高维正常混合物的自然延伸。然而,关键的区别在于,均值函数和协方差函数都采用了平滑结构。将期望最大化(EM)算法、核回归和函数主成分分析的思想结合起来,证明了该模型是可辨识的,并且可以有效地估计。我们的方法通过蒙特卡罗模拟得到了经验证明,并通过对超市数据集的分析进行了说明。
When functional data are not homogenous, for example, when there are multiple classes of functional curves in the dataset, traditional estimation methods may fail. In this article, we propose a new estimation procedure for the mixture of Gaussian processes, to incorporate both functional and inhomogenous properties of the data. Our method can be viewed as a natural extension of high-dimensional normal mixtures. However, the key difference is that smoothed structures are imposed for both the mean and covariance functions. The model is shown to be identifiable, and can be estimated efficiently by a combination of the ideas from expectation-maximization (EM) algorithm, kernel regression, and functional principal component analysis. Our methodology is empirically justified by Monte Carlo simulations and illustrated by an analysis of a supermarket dataset.
DOI: 10.1214/009053606000000272
发表时间: 2006-06-01
影响因子: 4.5
作者:
Hall, Peter;Mueller, Hans-Georg;Wang, Jane-Ling
通讯作者: Wang, Jane-Ling
DOI: 10.1198/016214505000000187
发表时间: 2006-03-01
影响因子: 3.7
作者:
Heard, NA;Holmes, CC;Stephens, DA
通讯作者: Stephens, DA
DOI: 10.1214/aoms/1177698520
发表时间: 1968-01-01
影响因子: --
作者:
YAKOWITZ, SJ;SPRAGINS, JD
通讯作者: SPRAGINS, JD
DOI: 10.1111/j.1541-0420.2007.00758.x
发表时间: 2007-09-01
期刊: BIOMETRICS
影响因子: 1.9
作者:
Shi, J. Q.;Wang, B.;Titterington, D. M.
通讯作者: Titterington, D. M.
DOI: 10.1093/bioinformatics/btg014
发表时间: 2003-03-01
期刊: BIOINFORMATICS
影响因子: 5.8
作者:
Luan, YH;Li, HZ
通讯作者: Li, HZ