Bayesian Estimation for the Extended t-process Regression Models with Independent Errors

Abstract

The extended t-process regression model is developed to robustly model functional data with outlier functional curves. This paper applies Bayesian estimation to propose an estimation procedure for the model with independent errors. A Monte Carlo EM method is built to estimate parameters involved in the model. Simulation studies and real examples show the proposed method performs well against outliers.

Keywords

Extended t-process regression / Functional data / Robustness / Monte Carlo EM algorithm

References

[1] Copt S, Victoria-Feser MP. High Breakdown Inference in the Mixed Linear Model. 2003 Rochester: Social Science Electronic Publishing

[2] Dempster AP, Laird NM, Rubin DB. Maximum likelihood from incomplete data via the em algorithm. J. R. Stat. Soc. Ser. B (Methodological). 1977, 39 1 1-22

[3] Fernandez C, Steel MFJ. Multivariate student-t regression model: Pitfalls and inferences. Biometrika. 1999, 86 1 153-167

[4] Lange K, Sinsheimer JS. Normal/independent distributions and their applications in robust regression. J. Comput. Graph. Stat.. 1993, 2 2 175-198

[5] Liu, C.: Statistical Analysis Using the Multivariate t Distribution. Ph.D. thesis, Harvard University (1994)

[6] Meza C, Osorio F, De la Cruz R. Estimation in nonlinear mixed-effects models using heavy-tailed distributions. Stat. Comput.. 2012, 22 1 121-139

[7] Neal, R.M.: Monte Carlo Implementation of Gaussian Process Models for Bayesian Regression and Classification. Technical report, Department of Statistics, University of Toronto (1997)

[8] Ramsay JO, Silverman BW. Functional Data Analysis. 2006 Berlin: Springer

[9] Rasmussen, C.E.: Evaluation of Gaussian Processes and Other Methods for Non-linear Regression. Ph.D. thesis, University of Toronto (1996)

[10] Rasmussen CE, Williams CKI. Gaussian Processes for Machine Learning. 2006 Cambridge: The MIT Press

[11] Relles DA, Rogers WH. Statisticians are fairly robust estimators of location. J. Am. Stat. Assoc.. 1977, 72 357 107-111

[12] Shi JQ, Choi T. Gaussian Process Regression Analysis for Functional Data. 2011 London: Chapman and Hall/CRC

[13] Wang, Z., Li, K., Shi, J.Q.: A Robust Estimation for the Extended t-Process Regression Model. (2018a). arXiv:1812.07701

[14] Wang, Z., Noh, M., Lee, Y., Shi, J.Q.: A Robust t-Process Regression Model with Independent Errors. (2018b). arXiv:1707.02014

[15] Wang Z, Shi JQ, Lee Y. Extended t-process regression models. J. Stat. Plan. Inference. 2017, 189 38-60

[16] Wei G, Tanner M. A monte carlo implementation of the em algorithm and the poor mans data augmentation algorithms. J. Am. Stat. Assoc.. 1990, 85 699-704

[17] Xu P, Lee Y, Shi JQ, Eyre J. Automatic detection of significant areas for functional data with directional error control. Stat. Med.. 2019, 38 3 376-397

Funding

Anhui Provincial Natural Science Foundation(1908085MA06)

National Natural Science Foundation of China(11471302)