Consistency and asymptotic normality for a nonparametric prediction under measurement errors

Mynbaev, Kairat and Martins-Filho, Carlos (2015): Consistency and asymptotic normality for a nonparametric prediction under measurement errors. Published in: Journal of Multivariate Analysis , Vol. 139, (2015): pp. 166-188.

Abstract

Nonparametric prediction of a random variable Y conditional on the value of an explanatory variable X is a classical and important problem in Statistics. The problem is significantly complicated if there are heterogeneously distributed measurement errors on the observed values of X used in estimation and prediction. Carroll et al. (2009) have recently proposed a kernel deconvolution estimator and obtained its consistency. In this paper we use the kernels proposed in Mynbaev and Martins-Filho (2010) to define a class of deconvolution estimators for prediction that contains their estimator as one of its elements. First, we obtain consistency of the estimators under much less restrictive conditions. Specifically, contrary to what is routinely assumed in the extant literature, the Fourier transform of the underlying kernels is not required to have compact support, higher-order restrictions on the kernel can be avoided and fractional smoothness of the involved densities is allowed. Second, we obtain asymptotic normality of the estimators under the assumption that there are two types of measurement errors on the observed values of X. It is apparent from our study that even in this simplified setting there are multiple cases exhibiting different asymptotic behavior. Our proof focuses on the case where measurement errors are super-smooth and we use it to discuss other possibilities. The results of a Monte Carlo simulation are provided to compare the performance of the estimator using traditional kernels and those proposed in Mynbaev and Martins-Filho (2010).

• Item Type: MPRA Paper
• Language: English
• Keywords: Measurement errors, nonparametric prediction, asymptotic normality, Lipschitz conditions
• Subjects: C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General > C14 - Semiparametric and Nonparametric Methods: General

References

[1] R.A. Adams, J. Fournier, Sobolev Spaces, second ed., in: Pure and Applied Mathematics, vol. 140, Elsevier Science B.V, Amsterdam, 2003.

[2] D.G. Altman, J.M. Bland, Measurement in medicine: the analysis of method comparison studies, The Statistician 32 (1983) 307–317.

[3] O. Besov, V. Ilin, S. Nikolski, Integral Representations of Functions and Embedding Theorems, Wiley, New York, NY, 1978.

[4] P. Billingsley, Probability and Measure, third ed., John Wiley and Sons, 1995.

[5] R.J. Carroll, A. Delaigle, P. Hall, Nonparametric prediction in measurement error models, J. Amer. Statist. Assoc. 104 (2009) 993–1014.

[6] R.J. Carroll, D. Ruppert, L. Stefanski, C. Crainiceau, Measurement error in nonlinear models, in: Monographs in Statistics and Applied Probability, vol. 105, Chapman and Hall, New York, NY, 2006.

[7] E.A. Cator, Deconvolution with arbitrarily smooth kernels, Statist. Probab. Lett. 54 (2001) 205–214.

[8] S.X. Chen, Probability density function estimation using gamma kernels, Ann. Inst. Statist. Math. 52 (2000) 471–480.

[9] C.-L. Cheng, J. Riu, On estimating linear relationships when both variables are subject to heteroscedastic measurement errors, Technometrics 48 (2006) 511–518.

[10] A. Delaigle, J. Fan, R.J. Carroll, A design-adaptive local polynomial estimator for the errors-in-variables problem, J. Amer. Statist. Assoc. 104 (2009) 348–359.

[11] A. Delaigle, A. Meister, Nonparametric regression estimation in the heteroscedastic errors-in-variables problem, J. Amer. Statist. Assoc. 102 (2007) 1416–1426.

[12] A. Delaigle, A. Meister, Density estimation with heteroscedastic error, Bernoulli 14 (2008) 562–579.

[13] J. Fan, Asymptotic normality for deconvolution kernel density estimators, Sankhya 53 (1991) 97–110.

[14] J. Fan, Deconvolution with supersmooth distributions, Canad. J. Statist. 20 (1992) 155–169.

[15] Y. Fan, Y. Liu, A Note on asymptotic normality for deconvolution kernel density estimators, Sankhya 59 (1997) 138–141.

[16] J. Fan, E. Masry, Multivariate regression estimation with errors-in-variables: asymptotic normality for mixing processes, J. Multivariate Anal. 43 (1992) 237–271.

[17] J. Fan, Y. Truong, Nonparametric regression with errors in variables, Annals of Statistics 21 (1993) 1900–1925.

[18] J. Fan, Y. Truong, Y. Wang, Measurements Error Regression: A Nonparametric Approach. Tech. Rep. 2042, University of North Carolina - Institute of Statistics, 1990.

[19] T. Honda, Nonparametric regression for dependent data in the errors-in-variables problem, J. Statist. Plann. Inference 140 (2010) 3409–3424.

[20] S.B. Kulathinal, K. Kuulasmaa, D. Gasbarra, Estimation of errors-in-variables regression model when the variances of the measurement errors vary between the observations, Statistics in Medicine 21 (2002) 1089–1101.

[21] E. Masry, Asymptotic normality for deconvolution estimators of multivariate densities of stationary processes, J. Multivariate Anal. 44 (1993) 47–68.

[22] A. Meister, Deconvolution problems in nonparametric statistics, in: Lecture Notes in Statistics, vol. 193, Springer Verlag, 2009.

[23] K. Mynbaev, Distributions escaping to infinity and the limiting power of the Cliff-Ord test for autocorrelation, ISRN Probab. Stat. (2012) 1–39. Article number 926164.

[24] K. Mynbaev, C. Martins-Filho, Bias reduction in kernel density estimation via Lipschitz condition, J. Nonparametr. Stat. 22 (2010) 219–235.

[25] S.M. Nikolskii, Priblizhenie Funktsii Mnogikh Peremennykh i Teoremy Vlozheniya, second ed., Nauka, Moscow, 1977.

[26] J. Staudenmayer, D. Ruppert, J. Buonnacorsi, Density estimation in the presence of heteroskedastic measurement error, J. Amer. Statist. Assoc. 103 (2008) 726–736.

[27] H. Triebel, Theory of Function Spaces, Birkhauser Verlag, Basel, 1983.

[28] A.J. van Es, H.-W. Uh, Asymptotic normality of nonparametric kernel type deconvolution density estimators: crossing the Cauchy boundary, J. Nonparametr. Stat. 16 (2004) 261–277.

[29] B. van Es, H.-W. Uh, Asymptotic normality of kernel-type deconvolution estimators, Scand. J. Stat. 32 (2005) 467–483.

[30] S.D. Walter, Variation in baseline risk as an explanation of heterogeneity in meta-analysis, Stat. Med. 16 (1997) 2883–2900.

[31] C.-H. Zhang, Fourier methods for estimating mixing densities and distributions, Ann. Statist. 18 (1990) 806–831.