By Topic

Can Ensemble Method Convert a 'Weak' Evolutionary Algorithm to a 'Strong' One?

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Shude Zhou ; Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing ; Zengqi Sun

The contribution of the paper is bringing ensemble method to the field of evolutionary computation. The conceptive model of evolutionary algorithm ensemble is introduced, in which a collection of evolutionary algorithms are designed to solve the same problem and each interact with others. Two implementation methods are invented: data-based ensemble and model-based ensemble. In data-based ensemble, componential evolutionary algorithm shares a common data pool with others, and population of each algorithm is sampled from the pool using bagging method. In model-based ensemble, there are a collection of models describing the evolution status, and they cooperate by the way of information interaction. As examples, simple genetic algorithm and PBIL (population based incremental learning) are used to implement the ideas respectively. Experiments on combinatorial optimization problems show that ensemble method improves the performance of evolutionary algorithm. It can be concluded ensemble method can convert a `weak' evolutionary algorithm to a `strong' one

Published in:

Computational Intelligence for Modelling, Control and Automation, 2005 and International Conference on Intelligent Agents, Web Technologies and Internet Commerce, International Conference on  (Volume:2 )

Date of Conference:

28-30 Nov. 2005