Title :
A comparative study of linear encoding in Genetic Programming
Author :
Suttasupa, Yuttana ; Rungraungsilp, Suppat ; Pinyopan, Suwat ; Wungchusunti, Pravit ; Chongstitvatana, Prabhas
Author_Institution :
Dept. of Comput. Eng., Chulalongkorn Univ., Bangkok, Thailand
Abstract :
Genetic Programming is a widely used technique to solve many optimization problems. The original representation of a solution is a tree structure. To improve its search capability there are many proposals for encoding data structure of a solution of Genetic Programming as a linear code. However there are a few work in comparing between these proposals. This work presents a systematic way to compare three popular techniques for linear encoding in Genetic Programming. They are Linear Genetic Programming, Gene Expression Programming and Multi-Expression Programming. Ten problems in Symbolic Expressions are defined and are used as benchmarks to compare the effectiveness of these proposals against the baseline standard Genetic Programming. The metrics of comparison are the Success Rate and the absolute error. The discussion and comparison of the strength and weakness of each method are also presented.
Keywords :
genetic algorithms; linear codes; linear programming; tree data structures; absolute error; data structure encoding; gene expression programming; linear code; linear encoding; linear genetic programming; multiexpression programming; optimization problem; success rate; symbolic expressions; tree structure; Biological cells; Computers; Encoding; Gene expression; Genetic algorithms; Genetic programming; Programming; Gene Expression Programming; Genetic programming; Linear Genetic Programming; Multi-expression Programming;
Conference_Titel :
ICT and Knowledge Engineering (ICT & Knowledge Engineering), 2011 9th International Conference on
Conference_Location :
Bangkok
Print_ISBN :
978-1-4577-2161-8
DOI :
10.1109/ICTKE.2012.6152392
