Clustering Analysis of the Population in Db_SHADE Algorithm

Clustering Analysis of the Population in Db_SHADE Algorithm

This paper provides an analysis of the population clustering in a novel Success-History based Adaptive Differential Evolution algorithm with Distance based adaptation (Db_SHADE) in order to analyze the exploration and exploitation abilities of the algorithm. The comparison with the original SHADE a...

Full description

Bibliographic Details
Main Authors: Adam Viktorin, Roman Senkerik, Michal Pluhacek, Tomas Kadavy
Format: Article
Language:English
Published: Brno University of Technology 2018-06-01
Series:Mendel
Subjects:
Distance based parameter adaptation
SHADE
Differential evolution
DBSCAN
Online Access:https://mendel-journal.org/index.php/mendel/article/view/13
Description
Summary:This paper provides an analysis of the population clustering in a novel Success-History based Adaptive Differential Evolution algorithm with Distance based adaptation (Db_SHADE) in order to analyze the exploration and exploitation abilities of the algorithm. The comparison with the original SHADE algorithm is performed on the CEC2015 benchmark set in two dimensional settings (10D and 30D). The clustering analysis helps to answer the question about prolonged exploration phase of the Db_SHADE algorithm. Possible future research directions are drawn in the discussion and conclusion.
ISSN:1803-3814
2571-3701

Similar Items