An adjusted simulated annealing approach to particle swarm optimization: empirical performance in decision making

Authors:
Dae Sung Lee;Young Wook Seo;Kun Chang Lee
Affiliations:
SKK Business School, Sungkyunkwan University, Seoul, Republic of Korea;Software Engineering Center at NIPA, Seoul, Republic of Korea;SKK Business School, WCU Professor at Department of Interaction Science, Sungkyunkwan University, Seoul, Republic of Korea
Venue:
ACIIDS'11 Proceedings of the Third international conference on Intelligent information and database systems - Volume Part II
Year:
2011

Citing 6
Cited 0

Evolution strategies –A comprehensive introduction

Natural Computing: an international journal
Molecular docking with multi-objective Particle Swarm Optimization

Applied Soft Computing
A hybrid cooperative-comprehensive learning based PSO algorithm for image segmentation using multilevel thresholding

Expert Systems with Applications: An International Journal
An effective hybrid PSO-based algorithm for flow shop scheduling with limited buffers

Computers and Operations Research
Letters: An improved PSO-based ANN with simulated annealing technique

Neurocomputing
Comprehensive learning particle swarm optimizer for global optimization of multimodal functions

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Particle swarm optimization (PSO) is a novel population-based searching technique proposed as an alternative to genetic algorithm (GA). It has had wide applications in a variety of fields. We suggest a hybrid clustering algorithm, which applies the combination of conventional PSO and SA (Simulated Annealing) algorithm to the process of K-means clustering in order to solve the problem of premature convergence. In addition we develop an adjustment algorithm, which modifies the acceleration constants of PSO by comparison of global and local best position, and is applied to the mixture algorithm named as SA-PSO so as to minimize the search of unnecessary areas and enhance performance. We simulated and compared three algorithms (K-PSO, SA-PSO and Adjusted SA-PSO). The results demonstrated our new approach (Adjusted SA-PSO) had the most excellent performance in usefulness and reliability evaluation, which denotes fitness function and mean absolute error respectively.