Learning and interpolation are two extreme variants of the same problem, the object of which is to construct a function which is supposed to reasonably approximate an unknown function of which only a certain number of samples are known. These problems appear in various frameworks which go from the solution of equations to the partial derivatives, while passing by the modeling. In this paper, we present the Bayesian method of reconstructing or approximating functions of some proposed functions, based on the Reproducing Kernel Hilbert Space using Gaussian processes.

Gaussian processes, spline interpolation, reproducing kernel, Bayesian inference.

Received: September 25, 2021; Accepted: November 16, 2021; Published: December 22, 2021

How to cite this article: Komi Agbokou and Yaogan Mensah, Inference on the Reproducing Kernel Hilbert Spaces, Universal Journal of Mathematics and Mathematical Sciences 15 (2022), 11-29. DOI: 10.17654/2277141722002

This Open Access Article is Licensed under Creative Commons Attribution 4.0 International License

References:

[1] K. Agbokou and K. E. Gneyou, On the strong convergence of the hazard rate and its maximum risk point estimators in presence of censorship and functional explanatory covariate, Afr. Stat. 12(3) (2017), 1397-1416.
[2] K. Agbokou, K. E. Gneyou and E. Deme, Almost sure representations of the conditional hazard function and its maximum estimation under right-censoring and left-truncation, Far East Journal of Theoretical Statistics 54(2) (2018), 141-173.
[3] K. Agbokou and K. E. Gneyou, On the asymptotic distribution of non-parametric conditional quantile estimator under random censorship, Far East Journal of Theoretical Statistics 56(1) (2019a), 1-34.
[4] K. Agbokou and K. E. Gneyou, Asymptotic properties of the conditional hazard function and its maximum estimation under right-censoring and left-truncation, Jordan Journal of Mathematics and Statistics 12(3) (2019b), 351-374.
[5] J. Agler, F. B. Yeh and N. J. Young, Reproducing Kernel Spaces and Applications, Operator Theory: Advances and Applications, Vol. 143, Springer Basel AG, 2003.
[6] A. Aronszajn, Reproducing kernel Hilbert spaces, Am. Math. Soc. Trans. 68 (1950), 337-404.
[7] R. K. Beatson and G. N. Newsam, Fast evaluation of radial basis functions: moment based methods, SIAM J. Sci. Comput. 19(5) (1998), 1428-1449.
[8] S. Bergman, Uber die Entwicklung der harmonischen Funktionen der Ebene und des Raumes nach Orthogonalfunktionen, Thesis, Berlin, 1921.
[9] S. Bochner, Vorlesungen iiber Fouriersche Integrale, Akademische Verlag-Gesellschaft, Leipzig, 1932.
[10] A. Caponnetto and E. D. Vito, Optimal rates for regularized least-squares algorithm, Foundations of Computational Mathematics 7(4) (2007), 331-368.
[11] R. M. Dudley, Real Analysis and Probability, Cambridge University Press, 2002.
[12] F. Girosi, M. Jones and T. Poggio, Regularization theory and neural networks architectures, Neural Computation 7(2) (1995), 219-269.
[13] K. E. Gneyou, Hazard rate prediction in life time data analysis, Contemporary Problems in Mathematical Physics, World Sci. Publ., Hackensack, NJ, 2006, pp. 397-411.
[14] M. Kanagawa, P. Hennig, D. Sejdinovic and K. B. Sriperumbudur, Gaussian Processes and Kernel Methods: A Review on Connections and Equivalences, 2018. arXiv:1807.02582v1 [stat.ML] 07/10/2018.
[15] J. Mercer, Functions of positive and negative type and their connection with the theory of integral equations, Philos. Trans. Roy. Soc. London, A 209 (1909), 415-446.
[16] E. Parzen, Statistical inference on time series by Hilbert space methods I, Tech. rep. 23, Applied Mathematics and Statistics Laboratory, Stanford University, California, 1959.
[17] E. Parzen, Probability density functionals and reproducing kernel Hilbert spaces, M. Rosenblatt, ed., Proceedings of the Symposium on Time Series Analysis, Wiley, 1963, pp. 155-169.
[18] V. N. Vapnik, The support vector method, Lecture Notes in Computer Science, 1997, p. 1327.
[19] A. N. Tikhonov and V. I. Arsenin, Solutions of Ill Posed Problems, W. H. Winston and Sons, John Wiley and Sons, Washington DC, 1977.
[20] G. Wahba, Spline models for observational data, Number 59 in CBMS-NSF Regional Conferences Series in Applied Mathematics, SIAM, 1990.
[21] S. Zaremba, L’équation biharmonique et une classe remarquable de fonctions fondamentales harmoniques, Bulletin International de l’Académie des Sciences de Cracovie, 1907, pp. 147-196.