作者: Pei He , Zelin Deng , Houfeng Wang , Zhusong Liu
DOI: 10.1007/S00500-015-1710-9
关键词: Mathematical proof 、 Grammatical evolution 、 Modular design 、 Semantic analysis (machine learning) 、 Modularity 、 Parsing 、 Theoretical computer science 、 Computational intelligence 、 Semantic computing 、 Computer science 、 Genetic programming
摘要: Many deficiencies with grammatical evolution (GE) such as inconvenience in solution derivations, modularity analysis, and semantic computing can partly be explained from the angle of genotypic representations. In this paper, we deepen some our previous work visualizing concept relationships, individual structures total evolutionary process, contributing new ideas, perspectives, methods these aspects; reveal principle hidden early so that to develop a practical methodology; provide formal proofs for issues concern which will helpful understanding mathematical essence issues, establishing an unified framework well implementation; exploit like modular discovery systematically lack supporting mechanism, if not impossible, is done poorly many existing systems, finally demonstrate possible gains through analysis reuse. As shown work, search space number nodes parser tree are reduced using concepts building blocks, codon-to-grammar mapping integer modulo arithmetic used most GE abnegated.