Abstract
We describe computer-based experiments evaluating the significance of the control parameters of a mapping procedure which is an extension of most of the known linear mapping techniques. The guidelines for control parameter variations are found. By this means the number of trials needed in the search for high quality mapping is reduced to a great extent.
Original language | English |
---|---|
Pages (from-to) | 637-645 |
Number of pages | 9 |
Journal | Pattern Recognition Letters |
Volume | 14 |
Issue number | 8 |
DOIs | |
State | Published - 1 Jan 1993 |
Keywords
- Mapping of multidimensional data
- measures of association
- nearest neighbour technique
- nonparametric scatter matrices
ASJC Scopus subject areas
- Software
- Signal Processing
- Computer Vision and Pattern Recognition
- Artificial Intelligence