Abstract
By using the algebraic approach to pattern recognition problems and the theory of local and universal constraints, universal constraints are described as sets of mappings that preserve m-place predicates. It is shown that symmetric and functional constraints admit a similar description.
Original language | English |
---|---|
Pages (from-to) | 527-532 |
Number of pages | 6 |
Journal | Computational Mathematics and Mathematical Physics |
Volume | 47 |
Issue number | 3 |
DOIs | |
Publication status | Published - Mar 1 2007 |
Keywords
- Algebraic approach
- Pattern recognition
- Precedent training
- Theory of local and universal constraints
ASJC Scopus subject areas
- Computational Mathematics