By Topic

Improving the performance of particle swarms through dimension reductions — A case study with locust swarms

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)
Chen, S. ; Sch. of Inf. Technol., York Univ., Toronto, ON, Canada ; Vargas, Y.N.

A key challenge for many heuristic search techniques is scalability - techniques that work well on low-dimension problems may perform poorly on high-dimension problems. To the extent that some problems/problem domains are separable, this can lead to a benefit for search techniques that can exploit separability. The standard algorithm for particle swarm optimization does not provide opportunities to exploit separable problems. However, the design of locust swarms involves two phases (scouts and swarms), and “dimension reductions” can be easily implemented during the scouts phase. This ability to exploit separability in locust swarms leads to large performance improvements on separable problems. More interestingly, dimension reductions can also lead to significant performance improvements on non-separable problems. Results on the Black-Box Optimization Benchmarking (BBOB) problems show how dimension reductions can help locust swarms perform better than standard particle swarms - especially on high-dimension problems.

Published in:

Evolutionary Computation (CEC), 2010 IEEE Congress on

Date of Conference:

18-23 July 2010