AUTHORS: Mingxing Zhang, Zixin Liu

Download as PDF

ABSTRACT: This paper investigates a class of estimation problems of the semiparametric model with missing data. In order to overcome the robust defect of traditional complete data estimation method and regression imputation estimation technique, we propose a modified imputation estimation approach called Kriging-regression imputation. Compared with previous method used in the references cited therein , the new proposed method not only makes more use of the data information, but also has better robustness. Model estimation and asymptotic distribution of the estimators are also derived theoretically. In order to improve the robustness, LASSO technique is further introduced into Kriging-regression imputation. Numerical experiment is also provided to show the effectiveness and superiority of our method.

KEYWORDS: semiparameter model, data missing, imputation techniques, asymptotic normality, consistency

REFERENCES:

[1] W. J. Conover, Practical Nonparametric Statistics, John–Wiley– Sons–New York 1980

[2] P. J. Green, B. W. Silverman, Nonparametric Regression and Generalized Linear Models: A Roughness Penalty Approach, Chapman–Hall 1994

[3] H. Midi, M. A. Mohammed ,The Performance of Robust Latent Root Regression Based on MM and modified GM estimators, WSEAS. Math. Trans. 13, 2014, pp. 916-924.

[4] P. Pinto, L. Pinto, The Use of Parametric Statistical Tests to Detect Network Traffic Anomalies on SDN Architectures, WSEAS. Commu. Trans. 13, 2014, pp. 659-664.

[5] R. Adeogun, Multipath Parameter Estimation and Channel Prediction for Wideband Mobile to Mobile Wireless Channel. WSEAS. Commu. Trans. 13, 2014, pp. 201-208

[6] J. Q. Fan, I. Gijbels, Local Polynomial Modelling and Its Applications, Chapman–Hall 1986

[7] A. M. Mood, The Distribution Theory of Certain Nonparametric Two-sample Tests, Ann. Math. Statist. 25, 1954, pp. 367–392.

[8] T. J,Hastile, R. J.Tibshirani, Generalized Additive Models, Monographs on Statistics and Applied Probability, Chapman–Hall–London 1990

[9] Z. W. Cai, J. Q Fan, Q. W. Yao, Efficient Estimation and Inference for Varying Coefficient Model, J. Amer. Statist. Assoc. 95, 2000, pp. 888– 902.

[10] R. Chen, R. J. Tasy, Functional-coefficient Autoregressive Models. J. Amer. Statist. Assoc. 88, 1993, pp. 298–308.

[11] T. J,Hastile, R. J.Tibshirani, Varying-coefficient Models(with Discussion), J. Roy. Statist. Soc. Ser. B. 55,1993, pp. 757–796.

[12] J. Z. Huang, C. O. Wu, L. Zhou, Varyingcoefficient Models and Basis Function Approximations for the Analysis of Repeated Measurements, Biometrika. 89, 2002, pp. 111–128.

[13] J. Q. Fan, W. Y. Zhang, Statistical Estimation in Varying-coefficient Models, J. Roy. Statist. Soc. Ser. B. 27(5), 1999, pp. 1491–1581.

[14] Y. C. Xia, W. K. Li, On the Estimation and Testing of Functional-coefficient Linear Models, Statist. Sinica. 9, 1999, 737–757.

[15] D. R. Hoover, J. A. Rice, C. O. Wu, L. P.Yang, Nonparametric Smoothing Estimates of Timevarying Coefficient Models With Longitudinal Data. Biometrika.85(4), 1998, pp. 809–822.

[16] T. Hu, Y. C.Xia, Adaptive Semivarying Coefficient Model Selection, Statistica. Sinica 22, 2010, pp. 575-599.

[17] J. Z. Huang, C. O. Wu, L. Zhou, Polynominal Spline Estimation and Inference for Varying Coefficient Models with Longitudinal Data, Statistic Sinica. 14, 2004, pp. 763–788.

[18] R. F. Engle, W. J. Granger, J. Rice, A. Weiss, Semiparametric Estimates of the Relation Between Weather and Electricity Techniques, J. Amer. Statist. Assoc. 80, 1986, pp. 310–319.

[19] H. Chen, Convergence Rates for Parametric Components in a Partially Linear Model, Ann. Statist. 88, 1988, pp. 298–308.

[20] A. Yatchew, An Elementary Estimator of the Partial Linear Model, Economics. Lett. 57, 1997, pp. 135–143.

[21] H. Liang, Estimation in Partially Linear Models and Numerical Comparisons, Comput. Statist. Data. Anal. 50, 2006, pp. 675–687.

[22] P. Speckman, Kernel Smoothing in Partial Linear Models, J. Roy. Statist. Soc. Ser. B. 50, 1988, PP. 413–416.

[23] H. Liang, W. Hardle, R. J. Carroll, Estimation in a Semiparametric Partially Linear Errorsin-variables Model, Ann. Statist. 27, 1999, pp. 1519–1535.

[24] J. Fan, T. Huang, Profile Likelihood Inferences on Semiparametric Varying-coefficient Partially Linear Models, Bernoulli . 11, 2005, pp. 1031– 1057.

[25] X. Zhou, J. You, Wavelet Estimation in Varyingcoefficient Partially Linear Regression Models, Statist. Probab. Lett. 68, 2004, pp. 91–104.

[26] C. H. Wei, X. Z.Wu, Profile Lagrange Multiplier Test for Partially Linear Varying-coefficient Regression Models, Sys. Sci. Math. Scis. 28, 2008, pp. 416–424.

[27] W. Zhang, S. Y. Lee, X. Song, Local Polynomial Fitting in Semivarying Coefficient Models, J. Multi. Anal. 82, 2002, pp. 166–188.

[28] J. H. You, G. Chen, Estimation of a Semiparametric Varying-coefficient Partially Linear Errors-in-variables Model, J. Multivar. Anal. 97, 2004, pp. 324–341.

[29] C. H. Wei, Estimation in Partially Linear Varying-coefficient Errors-in-Variables Models with Missing Responses, Acta. Mathematic. Scientia. 30A(4), 2010, pp. 1042–1054.

[30] R. Little, D. Rubin, Statistical Analysis with Missing Data, Technometrics. 45(4), 2003, pp. 364–365.

[31] C. Chu, P. Cheng, Nonparametric Regression Estimation with Missing Data, J. Statist. Planning. Inference. 48, 1995, PP. 85–99.

[32] Q. Wang, O. Linton, W, Hardle, Semiparametric Regression Analysis with Missing Response at Random.J. Multivar. Anal. 98, 2007, pp. 334– 345.

[33] H. Zou, The Adaptive Lasso and Its Oracle Properties, J. Amer. Statist. Assoc. 2006, 101, pp.1418–1429.

[34] Y. P. Mack, B. W. Silverman, Weak and Strong Uniform Consistency of Kernel Regression Estimates, Z. Wahrsch. Verw. Gebiete. 61, 1982, pp. 405–415.