2003 | OriginalPaper | Buchkapitel
Self-Organizing Graph Edit Distance
verfasst von : Michel Neuhaus, Horst Bunke
Erschienen in: Graph Based Representations in Pattern Recognition
Verlag: Springer Berlin Heidelberg
Enthalten in: Professional Book Archive
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
This paper addresses the issue of learning graph edit distance cost functions for numerically labeled graphs from a corpus of sample graphs. We propose a system of self-organizing maps representing attribute distance spaces that encode edit operation costs. The self-organizing maps are iteratively adapted to minimize the edit distance of those graphs that are required to be similar. To demonstrate the learning effect, the distance model is applied to graphs representing line drawings and diatoms.