作者: Lawrence B. Holder , Jacek P. Kukluk , Diane J. Cook , Chang Hun You
DOI:
关键词:
摘要: This paper describes graph-based relational, unsupervised learning algorithm to infer node replacement graph grammar and its application metabolic pathways. We search for frequent subgraphs then check overlap among the instances of in input graph. If by one node, we propose a production. also can hierarchy productions compressing portions described production inferring new on compressed show curves how process changes when increase size sample set. examine computation time with an increased number nodes graphs. inferred grammars from pathways which do not change more graphs It indicates that found represent sets well.