We introduce flexible robust functional regression models, using various heavy-tailed processes, including a Student t-process. We propose efficient algorithms in estimating parameters for the marginal mean inferences and in predicting conditional means as well as interpolation and extrapolation for the subject-specific inferences. We develop bootstrap prediction intervals (PIs) for conditional mean curves. Numerical studies show that the proposed model provides a robust approach against data contamination or distribution misspecification, and the proposed PIs maintain the nominal confidence levels. A real data application is presented as an illustrative example.
RasmussenCEWilliamsCKI. Gaussian processes for machine learning, Cambridge, MA: MIT Press, 2006.
2.
ShiJQWangBMurray-SmithRet al.Gaussian process functional regression modelling for batch data. Biometrics2007; 63(3): 714–723.
3.
ShiJQWangBWillEJet al.Mixed-effects GPFR models with application to dose–response curve prediction. Stat Med2012; 31(26): 3165–3177.
4.
GramacyRLianH. Gaussian process single-index models as emulators for computer experiments. Technometrics2012; 54(1): 30–41.
5.
WangBShiJQ. Generalized Gaussian process regression model for non-Gaussian functional data. J Am Stat Assoc2014; 109(507): 1123–1133.
6.
AndrewsDFMallowsCL. Scale mixtures of normal distributions. J R Soc Series B Stat Methodol1974; 36(1): 99–102.
7.
LachosVHBandyopadhyayDDeyDK. Linear and nonlinear mixed-effects models for censored HIV viral loads using normal/independent distributions. Biometrics2011; 67(4): 1594–1604.
8.
MezaCOsorioFCruzR. Estimation in nonlinear mixed-effects models using heavy-tailed distributions. Stat Comput2012; 22(1): 121–139.
9.
CaoCZLinJGShiJQet al.Multivariate measurement error models for replicated data under heavy-tailed distributions. J Chemometr2015; 29(8): 457–466.
10.
BlasBBolfarineHLachosVH. Heavy tailed calibration model with Berkson measurement errors for replicated data. Chemometr Intell Lab2016; 156: 21–35.
11.
ZhuHBrownPJMorrisJS. Robust, adaptive functional regression in functional mixed model framework. J Am Stat Assoc2011; 106(495): 1167–1179.
12.
OsorioF. Influence diagnostics for robust P-splines using scale mixture of normal distributions. Ann Inst Stat Math2016; 68(3): 589–619.
13.
LeeYNelderJAPawitanY. Generalized linear models with random-effects, unified analysis via H-likelihood, London: Chapman and Hall, 2006.
14.
McCullochCENeuhausJM. Prediction of random effects in linear and generalized linear models under model misspecification. Biometrics2011; 67(1): 270–279.
15.
LeeYKimG. H-likelihood predictive intervals for unobservables. Int Stat Rev2016; 84(3): 487–505.
16.
DigglePJLiangKYZegerSL. Analysis of longitudinal data, New York: Oxford Univ. Press, 1996.
17.
LeeYNelderJA. Conditional and marginal models: Another view (with discussion). Stat Sci2004; 19(2): 219–238.
18.
RamsayJOSilvermanBW. Functional data analysis, 2nd ed. New York: Springer, 2005.
19.
ShiJQChoiT. Gaussian process regression analysis for functional data, London: Chapman and Hall, 2011.
20.
LangeKSinsheimerJS. Normal/independent distributions and their applications in robust regression. J Comput Graph Stat1993; 2(2): 175–198.
21.
PinheiroJCLiuCWuYN. Efficient algorithms for robust estimation in linear mixed-effect models using the multivariate t distribution. J Comput Graph Stat2001; 10(2): 249–276.
22.
SavalliCPaulaGACysneirosFJA. Assessment of variance components in elliptical linear mixed models. Stat Model2006; 6(1): 59–76.
23.
Wang Z, Shi JQ and Lee Y. Extended T-process regression models, 2016; arXiv: 1511.03402.
24.
WestRMHarrisKGilthorpeMSet al.Functional data analysis applied to a randomized controlled clinical trial in hemodialysis patients describes the variability of patient responses in the control of renal anemia. J Am Soc Nephrol2007; 18(8): 2371–2376.
25.
TolmanCRichardsonDBartlettCet al.Structured conversion from thrice weekly to weekly erythropoietic regimens using a computerized decision-support system: A randomized clinical study. J Am Soc Nephrol2005; 16(5): 1463–1470.
26.
WillEJRichardsonDTolmanCet al.Development and exploitation of a clinical decision support system for the management of renal anaemia. Nephrol Dial Transplant2007; 22(Suppl. 4): iv31–iv36.
27.
SchwarzG. Estimating the dimension of a model. Ann Stat1978; 6(2): 461–464.
28.
FangKTKotzSNgKW. Symmetrical multivariate and related distributions, London: Chapman and Hall, 1990.
29.
LeeYNelderJA. Hierarchical generalized linear models (with discussion). J R Soc B Stat Methodol1996; 58(4): 619–678.
30.
MengXLRubinDB. Maximum likelihood estimation via the ECM algorithm: A general framework. Biometrika1993; 80(2): 267–278.
31.
LiuCHRubinDB. The ECME algorithm: A simple extension of EM and ECM with faster monotone convergence. Biometrika1994; 81(4): 633–648.
32.
LangeKLittleRTaylorJ. Robust statistical modeling using the t distribution. J Am Stat Assoc1989; 84(408): 881–896.
33.
BjørnstadJF. Predictive likelihood: A review. Stat Sci1990; 5(1): 242–265.
34.
BjørnstadJF. On the generalization of the likelihood function and likelihood principle. J Am Stat Assoc1996; 91(434): 791–806.
35.
SeegerMWKakadeSMFosterDP. Information consistency of nonparametric Gaussian process methods. IEEE Trans Inform Theory2008; 54(5): 2376–2382.
36.
YaoFMüllerHGWangJL. Functional data analysis for sparse longitudinal data. J Am Stat Assoc2005; 100(470): 577–590.
37.
LiYHsingT. Uniform convergence rates for nonparametric regression and principal component analysis in functional/longitudinal data. Ann Stat2010; 38(6): 3321–3351.
38.
YiGShiJQChoiT. Penalized Gaussian process regression and classification for high-dimensional nonlinear data. Biometrics2011; 67(4): 1285–1294.
AntoniadisASapatinasT. Estimation and inference in functional mixed-effects models. Comput Stat Data Anal2007; 51(10): 4793–4813.
41.
WuHZhangJT. Nonparametric regression methods for longitudinal data analysis: Mixed-effects modeling approaches, New Jersey: Wiley, 2006.
42.
RuppertDWandMPCarrollRJ. Semiparametric regression, Cambridge: Cambridge University Press, 2003.
43.
FangXLiJWongWKet al.Detecting the violation of variance homogeneity in mixed models. Stat Methods Med Res2016; 25(6): 2506–2520.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
