Title :
Quantile Regression Based on Laplacian Manifold Regularization
Author_Institution :
Econ. & Manage. Sch., Wuhan Univ., Wuhan, China
Abstract :
In this paper we present a nonparametric version of a quantile estimator based on Laplacian manifold regularization, which can be obtained by solving a simple quadratic programming problem, and provide bounds on the quantile property and uniform convergence statements of our estimator. Experimental results show the competitiveness of our method with existing ones.
Keywords :
Laplace equations; convergence; quadratic programming; regression analysis; Laplacian manifold regularization; nonparametric version; quadratic programming problem; quantile estimator; quantile property; quantile regression; uniform convergence statements; Algorithm design and analysis; Convergence; Estimation; Laplace equations; Manifolds; Optimization; Semisupervised learning; Laplacian Manifold Regularization; Nonlinear Quantile Regression; Semi-supervised Learning;
Conference_Titel :
Information Science and Control Engineering (ICISCE), 2015 2nd International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4673-6849-0
DOI :
10.1109/ICISCE.2015.92
