Robust functional principal components for sparse longitudinal data

被引:7
|
作者
Boente, Graciela [1 ,2 ,3 ]
Salibian-Barrera, Matias [4 ]
机构
[1] Univ Buenos Aires, Dept Matemat, Buenos Aires, DF, Argentina
[2] Univ Buenos Aires, Inst Calculo, Buenos Aires, DF, Argentina
[3] Consejo Nacl Invest Cient & Tecn, Buenos Aires, DF, Argentina
[4] Univ British Columbia, Dept Stat, Fac Sci, Vancouver, BC, Canada
来源
基金
加拿大自然科学与工程研究理事会;
关键词
Functional data analysis; Principal components; Robust estimation; Sparse data; NONPARAMETRIC REGRESSION; ELLIPTIC DISTRIBUTIONS; CONVERGENCE-RATES; ESTIMATORS; MODELS; EFFICIENT; LOCATION;
D O I
10.1007/s40300-020-00193-3
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
In this paper we review existing methods for robust functional principal component analysis (FPCA) and propose a new method for FPCA that can be applied to longitudinal data where only a few observations per trajectory are available. This method is robust against the presence of atypical observations, and can also be used to derive a new non-robust FPCA approach for sparsely observed functional data. We use local regression to estimate the values of the covariance function, taking advantage of the fact that for elliptically distributed random vectors the conditional location parameter of some of its components given others is a linear function of the conditioning set. This observation allows us to obtain robust FPCA estimators by using robust local regression methods. The finite sample performance of our proposal is explored through a simulation study that shows that, as expected, the robust method outperforms existing alternatives when the data are contaminated. Furthermore, we also see that for samples that do not contain outliers the non-robust variant of our proposal compares favourably to the existing alternative in the literature. A real data example is also presented.
引用
收藏
页码:159 / 188
页数:30
相关论文
共 50 条
  • [1] Robust functional principal components for sparse longitudinal data
    Graciela Boente
    Matías Salibián-Barrera
    [J]. METRON, 2021, 79 : 159 - 188
  • [2] Robust functional principal components for irregularly spaced longitudinal data
    Ricardo A. Maronna
    [J]. Statistical Papers, 2021, 62 : 1563 - 1582
  • [3] Robust functional principal components for irregularly spaced longitudinal data
    Maronna, Ricardo A.
    [J]. STATISTICAL PAPERS, 2021, 62 (04) : 1563 - 1582
  • [4] A Geometric Approach to Maximum Likelihood Estimation of the Functional Principal Components From Sparse Longitudinal Data
    Peng, Jie
    Paul, Debashis
    [J]. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 2009, 18 (04) : 995 - 1015
  • [5] BAYESIAN MULTIVARIATE SPARSE FUNCTIONAL PRINCIPAL COMPONENTS ANALYSIS WITH APPLICATION TO LONGITUDINAL MICROBIOME MULTIOMICS DATA
    Jiang, Lingjing
    Elrod, Chris
    Kim, Jane J.
    Swafford, Austin D.
    Knight, Rob
    Thompson, Wesley K.
    [J]. ANNALS OF APPLIED STATISTICS, 2022, 16 (04): : 2231 - 2249
  • [6] SPARSE AND FUNCTIONAL PRINCIPAL COMPONENTS ANALYSIS
    Allen, Genevera I.
    Weylandt, Michael
    [J]. 2019 IEEE DATA SCIENCE WORKSHOP (DSW), 2019, : 11 - 16
  • [7] Learning Robust and Sparse Principal Components With the α-Divergence
    Rekavandi, Aref Miri
    Seghouane, Abd-Krim
    Evans, Robin J.
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2024, 33 : 3441 - 3455
  • [8] COVARIATE ADJUSTED FUNCTIONAL PRINCIPAL COMPONENTS ANALYSIS FOR LONGITUDINAL DATA
    Jiang, Ci-Ren
    Wang, Jane-Ling
    [J]. ANNALS OF STATISTICS, 2010, 38 (02): : 1194 - 1226
  • [9] Joint modelling of paired sparse functional data using principal components
    Zhou, Lan
    Huang, Jianhua Z.
    Carroll, Raymond J.
    [J]. BIOMETRIKA, 2008, 95 (03) : 601 - 619
  • [10] DERIVATIVE PRINCIPAL COMPONENTS FOR REPRESENTING THE TIME DYNAMICS OF LONGITUDINAL AND FUNCTIONAL DATA
    Dai, Xiongtao
    Muller, Hans-Georg
    Tao, Wenwen
    [J]. STATISTICA SINICA, 2022, 32 (01) : 179 - 180