Optimal representation of multivariate functions or data in visualizable low-dimensional spaces |
| |
Authors: | Jian Song |
| |
Institution: | (1)Chinese Academy of Engineering ,Beijing 100038 ,China |
| |
Abstract: | It is intended to find the best representation of high-dimensional functions or multivariate data in L2(Ω ) with fewest number of terms, each of them is a combination of one-variable function. A system of nonlinear integral equations has been derived as an eigenvalue problem of gradient operator in the said space. It proved that the complete set of eigenfunctions generated by the gradient operator constitutes an orthonormal system, and any function of L2(Ω ) can be expanded with fewest terms and exponential rapidity of convergence. It is also proved as a corollary, the greatest eigenvalue of the integral operators has multiplicity 1 if the dimension of the underlying space Rn, n = 2, 4 and 6. |
| |
Keywords: | nonlinear integral equations gradient operators eigenvalues orthonormal system of eigenfunctions optimal approximation |
本文献已被 SpringerLink 等数据库收录! |
| 点击此处可从《中国科学通报(英文版)》浏览原始摘要信息 |
| 点击此处可从《中国科学通报(英文版)》下载免费的PDF全文 |