Journal of Computational and Applied Mathematics ( IF 2.4 ) Pub Date : 2020-06-13 , DOI: 10.1016/j.cam.2020.113061 Kerstin Hesse , Ian H. Sloan , Robert S. Womersley
In this paper we derive local error estimates for penalized least-squares approximation on the -dimensional unit sphere , given noisy, scattered, local data representing an underlying function from a Sobolev space of order defined on a non-empty connected open set with Lipschitz-continuous boundary. The quadratic regularization functional has two terms, one measuring the squared pointwise -discrepancy from the local data, the other containing the squared native space norm of a radial basis function (RBF), multiplied by a regularization parameter. The RBF is chosen so that its native space is equivalent to the (global) Sobolev space of order on . While both the data and the approximated function are local, we minimize the quadratic functional over all functions in the native space of the RBF, and obtain as exact minimizer a (global) radial basis function approximation. By choosing the RBF to be a Wendland function the resulting linear system has a sparse matrix which is easily computed. We consider three different strategies for choosing the smoothing parameter, namely Morozov’s discrepancy principle and two a priori strategies, and derive error estimates for each strategy. As auxiliary tools for proving the local error estimates we develop both a local sampling inequality and a suitable Sobolev extension theorem. The paper concludes with numerical experiments.
中文翻译:
含噪声分散数据的球面上基于局部RBF的惩罚最小二乘逼近
在本文中,我们推导了局部 罚最小二乘近似的误差估计 维单位球 ,给出嘈杂,分散的局部数据,这些数据表示Sobolev阶空间中的基础函数 在非空连接的开放集上定义 与Lipschitz连续边界。二次正则化函数具有两项,一项测量点平方的平方-与本地数据的差异,另一个包含径向基函数(RBF)的平方本机空间范数乘以正则化参数。选择RBF的目的是使其本机空间等于有序的(全局)Sobolev空间 上 。虽然数据和近似函数都是局部函数,但我们将RBF原始空间中所有函数的二次函数最小化,并获得(全局)径向基函数近似作为精确的最小化子。通过将RBF选择为Wendland函数,所得线性系统具有易于计算的稀疏矩阵。我们考虑选择平滑参数的三种不同策略,即莫罗佐夫的差异原理和两种先验策略,然后得出每个策略的错误估计。作为证明当地情况的辅助工具 误差估计我们都开发了本地 采样不等式和合适的Sobolev扩展定理。本文以数值实验作为结束。