On the generalization ability of GRLVQ networks

Autor(en): Hammer, B
Strickert, M
Villmann, T
Stichwörter: adaptive metric; Computer Science; Computer Science, Artificial Intelligence; generalization bounds; LVQ; margin optimization; relevance LVQ; SELECTION; SELF-ORGANIZING MAPS; VECTOR
Erscheinungsdatum: 2005
Herausgeber: SPRINGER
Journal: NEURAL PROCESSING LETTERS
Volumen: 21
Ausgabe: 2
Startseite: 109
Seitenende: 120
Zusammenfassung: 
We derive a generalization bound for prototype-based classifiers with adaptive metric. The bound depends on the margin of the classifier and is independent of the dimensionality of the data. It holds for classifiers based on the Euclidean metric extended by adaptive relevance terms. In particular, the result holds for relevance learning vector quantization (RLVQ) [4] and generalized relevance learning vector quantization (GRLVQ) [19].
ISSN: 13704621
DOI: 10.1007/s11063-004-1547-1

Show full item record

Google ScholarTM

Check

Altmetric