Latent function-on-scalar regression models for observed sequences of binary data: a restricted likelihood approach

In this paper, we study a functional regression setting where the random response curve is unobserved, and only its dichotomized version observed at a sequence of correlated binary data is available. We propose a practical computational framework for maximum likelihood analysis via the parameter expansion technique. Compared to existing methods, our proposal relies on the use of a complete data likelihood, with the advantage of being able to handle non-equally spaced and missing observations effectively. The proposed method is used in the Function-on-Scalar regression setting, with the latent response variable being a Gaussian random element taking values in a separable Hilbert space. Smooth estimations of functional regression coefficients and principal components are provided by introducing an adaptive MCEM algorithm that circumvents selecting the smoothing parameters. Finally, the performance of our novel method is demonstrated by various simulation studies and on a real case study. The proposed method is implemented in the R package dfrr.

[1]  T. Hsing,et al.  Uniform convergence rates for nonparametric regression and principal component analysis in functional/longitudinal data , 2010, 1211.2137.

[2]  Patrick J Heagerty,et al.  Marginalized Models for Moderate to Long Series of Longitudinal Binary Response Data , 2007, Biometrics.

[3]  Catherine A. Sugar,et al.  Principal component models for sparse functional data , 1999 .

[4]  Xin Zhao,et al.  The functional data analysis view of longitudinal data , 2004 .

[5]  H. Müller,et al.  Modelling sparse generalized longitudinal observations with latent Gaussian processes , 2008 .

[6]  D. Rubin,et al.  Parameter expansion to accelerate EM : The PX-EM algorithm , 1997 .

[7]  Hans-Georg Müller,et al.  Functional Data Analysis , 2016 .

[8]  Bo Wang,et al.  Generalized Gaussian Process Regression Model for Non-Gaussian Functional Data , 2014, 1401.8189.

[9]  H. Müller,et al.  Functional Data Analysis for Sparse Longitudinal Data , 2005 .

[10]  T. Hsing,et al.  Theoretical foundations of functional data analysis, with an introduction to linear operators , 2015 .

[11]  Catherine A. Sugar,et al.  Clustering for Sparsely Sampled Functional Data , 2003 .

[12]  Jeffrey S. Morris,et al.  Bayesian function‐on‐function regression for multilevel functional data , 2015, Biometrics.

[13]  P. Diggle.  Analysis of Longitudinal Data , 1995 .

[14]  Hua Liang,et al.  Efficient Estimation of the Nonparametric Mean and Covariance Functions for Longitudinal and Sparse Functional Data , 2018, Journal of the American Statistical Association.

[15]  Chih-Li Sung,et al.  A Generalized Gaussian Process Model for Computer Experiments With Binary Time Series , 2017, Journal of the American Statistical Association.

[16]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[17]  A. Linde.  A Bayesian latent variable approach to functional principal components analysis with binary and count data , 2009 .

[18]  Hans-Georg Ller,et al.  Functional Modelling and Classification of Longitudinal Data. , 2005 .

[19]  J. Schrack,et al.  Generalized multilevel function‐on‐scalar regression and principal component analysis , 2015, Biometrics.

[20]  Raymond J Carroll,et al.  Hierarchical functional data with mixed continuous and binary measurements , 2014, Biometrics.

[21]  Jun S. Liu,et al.  Parameter Expansion for Data Augmentation , 1999 .

[22]  Huiping Xu,et al.  Likelihood Analysis of Multivariate Probit Models Using a Parameter Expanded MCEM Algorithm , 2010, Technometrics.

[23]  Fabian Scheipl,et al.  Generalized Functional Additive Mixed Models , 2015, 1506.05384.