Learning Structural Loss Parameters on Graph Embedding Applied on Symbolic Graphs

Producció científica: Informe/llibreLlibre d'ActesRecercaAvaluat per experts

1 Citació (Scopus)

Resum

We propose an amelioration of proposed Graph Embedding (GEM) method in previous work that takes advantages of structural pattern representation and the structured distortion. it models an Attributed Graph (AG) as a Probabilistic Graphical Model (PGM). Then, it learns the parameters of this PGM presented by a vector, as new signature of AG in a lower dimensional vectorial space. We focus to adapt the structured learning algorithm via 1-slack formulation with a suitable risk function, called Graph Edit Distance (GED). It defines the dissimilarity of the ground truth and predicted graph labels. It determines by the error tolerant graph matching using bipartite graph matching algorithm. We apply Structured Support Vector Machines (SSVM) to process classification task. During our experiments, we got our results on the GREC dataset.

Idioma originalAnglès
Nombre de pàgines2
DOIs
Estat de la publicacióPublicada - 25 de gen. 2018

Sèrie de publicacions

NomProceedings of the International Conference on Document Analysis and Recognition, ICDAR
Volum2
ISSN (imprès)1520-5363

Fingerprint

Navegar pels temes de recerca de 'Learning Structural Loss Parameters on Graph Embedding Applied on Symbolic Graphs'. Junts formen un fingerprint únic.

Com citar-ho