The time complexity analysis of a class of gene expression programming

This paper studies the time complexity of gene expression programming based on maintaining elitist (ME-GEP). Using the theory of Markov chain and the technique of artificial fitness level, the properties of transition matrices of ME-GEP are analyzed. Based on the properties, the upper and lower boun...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Soft computing (Berlin, Germany) Germany), 2015-06, Vol.19 (6), p.1611-1625
Hauptverfasser: Du, Xin, Ni, Youcong, Xie, Datong, Yao, Xin, Ye, Peng, Xiao, Ruliang
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper studies the time complexity of gene expression programming based on maintaining elitist (ME-GEP). Using the theory of Markov chain and the technique of artificial fitness level, the properties of transition matrices of ME-GEP are analyzed. Based on the properties, the upper and lower bounds of the average time complexity of ME-GEP are obtained. Furthermore, the upper bound is estimated, which is determined by the parameters of ME-GEP algorithm. And the theoretical results acquired in this paper are used to analyze ME-GEP for solving function modeling and clustering problem. At last, a set of experiments are performed on these problems to illustrate the effectiveness of theoretical results. The results show that the upper bound of expected first hitting time can be used to direct the algorithm design of ME-GEP.
ISSN:1432-7643
1433-7479
DOI:10.1007/s00500-014-1551-y