A Radial Basis Function Approximation for Large Datasets
This work addresses the computational bottleneck of solving large linear systems in RBF approximation for engineering problems, but the improvements are incremental.
The paper introduces a new approach to Radial Basis Function approximation for large datasets that uses matrix symmetry and block partitioning, and presents experimental results on real datasets showing improved computational efficiency.
Approximation of scattered data is often a task in many engineering problems. The Radial Basis Function (RBF) approximation is appropriate for large scattered datasets in d-dimensional space. It is non-separable approximation, as it is based on a distance between two points. This method leads to a solution of overdetermined linear system of equations. In this paper a new approach to the RBF approximation of large datasets is introduced and experimental results for different real datasets and different RBFs are presented with respect to the accuracy of computation. The proposed approach uses symmetry of matrix and partitioning matrix into blocks.