On multivariate discrete least squares
The discrete least squares problem determine a (generally unique) function NEWLINENEWLINE\[NEWLINEf=\sum_{j\in\mathbb{N}_m} c^*_j f_j\in \text{span\,}{\mathcal F},\;{\mathcal F}= \{f_j:j\in \mathbb{N}_m\}, \text{ where }\; \mathbb{N}_m=\{1,2,\ldots,m\}NEWLINE\]NEWLINE NEWLINEwhich minimizes the square of the \(\ell^2\)-norm NEWLINE\[NEWLINE\sum_{i\in\mathbb{N}_m}\,\Biggl(\sum_{j\in\mathbb{N}_m}c_j f_j(x_i)- y_i\Biggl)^2NEWLINE\]NEWLINE over all vectors \(\{c_j:j\in \mathbb{N}_m\}\in \mathbb{R}^m\).NEWLINENEWLINE The authors ask ``what happens if the components of data \(X=\{x_i:i\in\mathbb{N}_m\}\) and \(s\in O\) (some neighborhood of \(\mathbb{R}^d\) containing \(X\)) are nearly the same.NEWLINENEWLINE This problem was satisfactorily solved in the univariate case in Section 6 of [\textit{Y. Ju Lee} and \textit{C. A. Micchelli}, ibid. 174, 148--181 (2013; Zbl 1282.41002)].NEWLINENEWLINE In this paper they treat the significantly more difficult multivariate case using an approach recently provided by the authors [J. Math. Anal. Appl. 427, No. 1, 74--87 (2015; Zbl 1331.41003)].
- On the least squares fit by radial functions to multidimensional scattered data
- The approximation power of moving least-squares
- Multivariate interpolation of arbitrarily spaced data by moving least squares methods
- Uniform approximation by discrete least squares polynomials
- scientific article; zbMATH DE number 2127857
- A framework for moving least squares method with total variation minimizing regularization
- A study on multivariate interpolation by increasingly flat kernel functions
- scientific article; zbMATH DE number 3348831 (Why is no real title available?)
- On collocation matrices for interpolation and approximation
This page was built for publication: On multivariate discrete least squares
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q320336)
