Nonparametric Density Estimation Using Flexible NURBS Modeling

Year 2025, Volume: 17 Issue: 2, 434 - 440, 30.12.2025

Mahmut Sami Erdoğan

https://doi.org/10.47000/tjmcs.1733086

Abstract

Nonparametric methods for density estimation provide flexible tools for modeling data distributions without assuming a specific parametric form. In this study, we propose a novel approach based on Non-Uniform Rational B-Splines (NURBS) to estimate probability density functions in a fully data-driven manner. The estimator is constructed by optimizing a set of control points and associated weights under constraints that ensure non-negativity and unit integral, guaranteeing that the resulting function is a valid density. Unlike classical polynomial-based estimators, the NURBS framework offers enhanced flexibility by accommodating non-uniform knot vectors and rational weighting, allowing for better adaptation to sharp features and multimodal structures. The performance of the proposed estimator is examined through simulations involving a wide range of distributional shapes, and its practical performance is demonstrated using real-world datasets. Comparative results indicate that the NURBS-based estimator provides competitive or superior accuracy compared to traditional Bernstein and Bézier-based alternatives, especially in complex distributional settings.

Keywords

NURBS , density function estimation , Bernstein polynomials , mean squared error

References

• Babu, G.J., Canty, A.J., Chaubey, Y.P., Application of Bernstein polynomials for smooth estimation of a distribution and density function, J. Statist. Plann. Inference, 105(2)(2002), 377–392.
• Belalia, M., Bouezmarni, T., Leblanc, A., Smooth conditional distribution estimators using Bernstein polynomials, Comput. Stat. Data Anal., 111(2017), 166–182.
• Erdoğan, M.S., Dişibüyük, Ç., Oruç, Ö.E., An alternative distribution function estimation method using rational Bernstein polynomials, J. Comput. Appl. Math., 353(2019), 232–242.
• Ghosal, S., Convergence rates for density estimation with Bernstein polynomials, Ann. Statist., 29(5)(2001), 1264–1280.
• Janosi, A., Steinbrunn,W., Pfisterer, M., & Detrano, R., Heart Disease Dataset, UCI Machine Learning Repository, 1988, https://archive.ics.uci.edu/ml/datasets/heart+disease.
• Johnson, S.G., The NLopt nonlinear-optimization package, https://github.com/stevengj/nlopt.
• Leblanc, A., Chung–Smirnov property for Bernstein estimators of distribution functions, J. Nonparametr. Stat., 21(2)(2009), 133–142.
• Leblanc, A., A bias-reduced approach to density estimation using Bernstein polynomials, J. Nonparametr. Stat., 22(4)(2010), 459–475.
• Leblanc, A., On estimating distribution functions using Bernstein polynomials, Ann. Inst. Stat. Math., 64(2012), 919–943.
• Marron, J.S., Wand, M.P., Exact mean integrated squared error, Ann. Statist., 20(2)(1992), 712–736.
• Marsh, D., Applied Geometry for Computer Graphics and CAD, Springer, 2006.
• Parzen, E., On estimation of a probability density function and mode, Ann. Math. Statist., 33(3)(1962), 1065–1076.
• Petrone, S., Wasserman, L., Consistency of Bernstein polynomial posteriors, J. R. Stat. Soc. Ser. B Stat. Methodol., 64(1)(2002), 79–100.
• R Core Team, R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria, 2025. https://www.R-project.org/.
• Rosenblatt, M., A central limit theorem and a strong mixing condition, Proc. Natl. Acad. Sci. USA, 42(1)(1956), 43–47.
• Silverman, B.W., Density Estimation for Statistics and Data Analysis, Springer, New York, 1986.
• Turnbull, B.C., Ghosh, S.K., Unimodal density estimation using Bernstein polynomials, Comput. Stat. Data Anal., 72(2014), 13–29.
• Vitale, R.A., A Bernstein Polynomial Approach to Density Function Estimation, in: Statistical Inference and Related Topics, Academic Press, 1975.
• Wand, M.P., Jones, M.C., Kernel Smoothing, CRC Press, 1994.
• Xue, L., Wang, J., Distribution function estimation by constrained polynomial spline regression, J. Nonparametr. Stat., 22(4)(2010), 443–457.