Title :
Genotype-phenotype with BNF mapping: An automatic approach of producing a computer program
Author :
Abdulwahhab, Rasha S.
Author_Institution :
Collage of Inf. Technol., Ahlia Univ., Manama, Bahrain
Abstract :
BNGA is a computer programs production that employs the engine of GA as a base for building its programs. BNGAs´ genotypes (solutions) have a list of integers that are used to develop the phenotypes (programs). These phenotypes are developed by using the grammatical syntax of the programming language. In this paper, BNGA is investigated and designed to study the symbolic regression and classification problems. Detailed analysis of these problems are examined and the results were clearly in favor of BNGA. All results were compared experimentally with other suggested techniques. Through these experiments, BNGA proved as a competitive approach that can be applied in practice to solve the problems of both computer science and engineering.
Keywords :
genetic algorithms; pattern classification; programming; programming languages; BNF mapping; BNGA; backus normal form with GA; classification problem; computer program production; computer science; engineering; genetic algorithm; genotype-phenotype; grammatical syntax; programming language; symbolic regression; Breast; Heart; Indium phosphide; Iris; Radio access networks;
Conference_Titel :
Evolving and Adaptive Intelligent Systems (EAIS), 2012 IEEE Conference on
Conference_Location :
Madrid
Print_ISBN :
978-1-4673-1728-3
Electronic_ISBN :
978-1-4673-1726-9
DOI :
10.1109/EAIS.2012.6232825