Nonparametric estimation of multivariate density and its derivative by dependent data using gamma kernels

L. A. Markovich

Fundamentalʹnaâ i prikladnaâ matematika, Tome 22 (2018) no. 3, pp. 145-177.

Résumé

We consider the nonparametric estimation of the multivariate probability density function and its partial derivative with a support on nonnegative axis by dependent data. We use the class of kernel estimators with asymmetric gamma kernel functions. The gamma kernels are nonnegative, they may change their shape depending on the position on the semi-axis and possess good boundary properties for a wide class of densities. Asymptotic estimates of the multivariate density and of its partial derivatives such as biases, variances, and covariances are derived. The optimal bandwidth of both estimates is obtained as a minimum of the mean integrated squared error (MISE) by dependent data with a strong mixing. Optimal convergence rates of the MISE both for the density and its derivative are found.

Bibliographie

[1] Aksoy H., “Use of gamma distribution in hydrological analysis”, Turkish J. Eng. Environ. Sci., 24 (2000), 419–428

[2] Andrews D. W. K., First order autoregressive processes and strong mixing, Cowles Foundation Discussion Paper 66, Yale University, 1983

[3] Bhattacharya P. K., “Estimation of a probability density function and its derivatives”, Indian J. Statist., A 29 (1967), 373–382 | MR | Zbl

[4] Bosq S., Nonparametric Statistics for Stochastic Processes. Estimation and Prediction, Springer, New York, 1996 | MR

[5] Bouezmarni T., Rombouts J. V. K., “Nonparametric density estimation for multivariate bounded data”, J. Statist. Planning Inference, 140:1 (2007), 139–152 | DOI | MR

[6] Bouezmarni T., Rombouts J. V. K., “Nonparametric density estimation for positive times series”, Comput. Statist. Data Analysis, 54:2 (2010), 245–261 | DOI | MR | Zbl

[7] Bouezmarni T., Scaillet O., “Consistency of asymmetric kernel density estimators and smoothed histograms with application to income data”, Econometric Theory, 21:2 (2005), 390–412 | DOI | MR | Zbl

[8] De Brabanter K., de Brabanter J., de Moor B., “Nonparametric derivative estimation”, Proc. of the 23rd Benelux Conf. on Artificial Intelligence, BNAIC (Gent, Belgium, 2011), 75–81

[9] Brown B. M., Chen S. X., “Beta-Bernstein smoothing for regression curves with compact support”, Scand. J. Statist., 26 (1999), 47–59 | DOI | MR | Zbl

[10] Chen S. X., “Probability density function estimation using gamma kernels”, Ann. Inst. Statist. Math., 54 (2000), 471–480 | DOI | MR

[11] Dobrovidov A. V., Koshkin G. M., “Data-based nonparametric signal filtration”, Austr. J. Statist., 40:1 (2011), 15–24

[12] Dobrovidov A. V., Koshkin G. M., “Regularized data-based nonparametric filtration of stochastic signals”, Proc. of the World Congress on Engineering, v. 1, 2011, 333–337

[13] Dobrovidov A. V., Koshkin G. M., Vasiliev V. A., Non-Parametric State Space Models, Kendrick Press, 2012 | MR | Zbl

[14] Dobrovidov A. V., Markovich L. A., “Data-driven bandwidth choice for gamma kernel estimates of density derivatives on the positive semi-axis”, IFAC Proc. Volumes, 46:11 (2013), 500–505 | DOI

[15] Dobrovidov A. V., Markovich L. A., “Nonparametric gamma kernel estimators of density derivatives on positive semi-axis”, IFAC Proc. Volumes, 46:9 (2013), 910–915 | DOI

[16] Funke B., Kawka R., “Nonparametric density estimation for multivariate bounded data using two non-negative multiplicative bias correction methods”, Comput. Statist. Data Analysis, 92 (2015), 148–162 | DOI | MR | Zbl

[17] Furman E., “On a multivariate gamma distribution”, Statist. Probab. Lett., 78 (2008), 2353–2360 | DOI | MR | Zbl

[18] Hall P., Wehrly T. E., “A geometrical method for removing edge effects from kernel-type nonparametric regression estimators”, J. Am. Statist. Assoc., 86 (1991), 665–672 | DOI | MR

[19] Hürlimann W., “Analytical evaluation of economic risk capital for portfolios of gamma risks”, Turkish J. Eng. Environ. Sci., 31 (2001), 107–122 | MR

[20] Igarashi G., Kakizawa Y., “Bias corrections for some asymmetric kernel estimators”, J. Statist. Planning Inference, 159 (2015), 37–63 | DOI | MR | Zbl

[21] Jones M. C., “Simple boundary correction for density estimation kernel”, Statist. Comput., 3 (1993), 135–146 | DOI | MR

[22] Kushnir A. F., “Asymptotically optimal tests for a regression problem of testing hypotheses”, Theory Probab. Its Appl., 13:4 (1967), 647–666 | DOI | MR

[23] Lejeune M., Sarda P., “Smooth estimators of distribution and density functions”, Comput. Statist. Data Analysis, 14 (1992), 457–471 | DOI | MR | Zbl

[24] Markovich L. A., “Inferences from optimal filtering equation”, Lithuan. Math. J., 55:3 (2015), 413–432 | DOI | MR | Zbl

[25] Markovich L. A., “Gamma kernel estimation of the density derivative on the positive semi-axis by dependent data”, REVSTAT-Statist. J. Statist. Portugal., 14:3 (2016), 327–348 | MR | Zbl

[26] Mathal A. M., Moschopoulos P. G., “A form of multivariate gamma distribution”, Ann. Inst. Statist. Math., 44 (1992), 97–106 | DOI | MR

[27] Müller H. G., “Smooth optimum kernel estimators near endpoints”, Biometrika, 78:3 (1991), 521–530 | DOI | MR | Zbl

[28] Nadarajah S., “Reliability for some bivariate gamma distributions”, Math. Probab. Eng., 2 (2005), 151–163 | DOI | MR | Zbl

[29] Nadarajah S., Gupta A. K., “Some bivariate gamma distributions”, Appl. Math. Lett., 19 (2006), 767–774 | DOI | MR | Zbl

[30] Nadaraya É. A., “On non-parametric estimates of density functions and regression curves”, Theory Probab. Its Appl., 10:1 (1964), 186–190 | DOI | MR

[31] Parzen E., “On estimation of a probability density function and mode”, Ann. Math. Statist., 33:3 (1962), 1065–1076 | DOI | MR | Zbl

[32] Rosenblatt M., “Remarks on some nonparametric estimates of a density function”, Ann. Math. Statist., 27:3 (1956), 832–837 | DOI | MR | Zbl

[33] Sasaki H., Hyvärinen A., Sugiyama M., “Clustering via mode seeking by direct estimation of the gradient of a log-density”, Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2014, Lect. Notes Comput. Sci., 8726, eds. Calders T., Esposito F., Hüllermeier E., Meo R., Springer, Berlin, 2014 | MR

[34] Scaillet O., “Density estimation using inverse and reciprocal inverse Gaussian kernels”, J. Nonparam. Statist., 16 (2004), 217–226 | DOI | MR | Zbl

[35] Schuster E. F., “Estimation of a probability function and its derivatives”, Ann. Math. Statist., 40 (1969), 1187–1195 | DOI | MR | Zbl

[36] Schuster E. F., “Incorporating support constraints into nonparametric estimators of densities”, Commun. Statist. Theory Methods, 14 (1985), 1123–1136 | DOI | MR | Zbl

[37] Silverman B. W., Density Estimation for Statistics and Data Analysis, Chapman and Hall, London, 1986 | MR | Zbl

[38] Tsypkin Ya. Z., “Optimality in adaptive control systems”, Uncertainty and Control, Lect. Notes Control Information Sci., 70, ed. J. Ackermann, Springer, Berlin, 1985, 153–214 | DOI

[39] Turlach B. A., Bandwidth Selection in Kernel Density Estimation: A Review, CORE and Inst. de Statistique, 1993

[40] Wand M. P., Jones M. C., Kernel Smoothing, Chapman and Hall, London, 1995 | MR | Zbl

[41] Zhang S., “A note on the performance of the gamma kernel estimators at the boundary”, Statist. Probab. Lett., 80 (2010), 548–557 | DOI | MR | Zbl