Fast evaluation of radial basis functions: I
โ Scribed by R.K. Beatson; G.N. Newsam
- Publisher
- Elsevier Science
- Year
- 1992
- Tongue
- English
- Weight
- 987 KB
- Volume
- 24
- Category
- Article
- ISSN
- 0898-1221
No coin nor oath required. For personal study only.
โฆ Synopsis
This paper describes some new tedmiques for the rapid evaluation and fitting of radial balfis functions. The techniques are based on the hierarchical and multipole expmmions recently introduced by several authors for the calculation of ma~-body potentials. Cmmider in paxticnlar the N term thin-plate spline, a(z) ---E~,ffil djqb(x -zj), where
The direct evaluation of s at a single extra point requires an extra O(N) operations. This paper shows that, with judicious use of series expaxmious, the incre~nental cost of evaluating a(x) to within precltdon e, can be cut to O(1 + IlogeD operations. In particular, if A is the interpolation matrix, ai,j = qb(=i -xj), the technique allows computation of the matrix-vector product Ad in O(N), rather than the previously required O(N 2) operations, and using only O(N) storage. Fast, storage-efficlent, computation of this matrix-vector product makes pre-conditioned conjugate~gradie~t methods very attractive as solvers of the interpolation equations, Ad --y, when N is large.
๐ SIMILAR VOLUMES
This paper provides the results of our study on automatic classification of mouse chromosomes. A radial basis function neural network was compared with a multi-layer perceptron and a probabilistic neural network. The networks were trained and tested with 3723 chromosomes presented to each network as